AD_________________ Award Number: W81XWH-04-1-0687 TITLE: MMP-8, a Breast Cancer Bone Metastasis Suppressor Gene PRINCIPAL INVESTIGATOR: Nagarajan Selvamurugan, Ph.D. CONTRACTING ORGANIZATION: University of Medicine and Dentistry of New Jersey Piscataway, NJ 08854 REPORT DATE: August 2006 TYPE OF REPORT: Final PREPARED FOR: U.S. Army Medical Research and Materiel Command Fort Detrick, Maryland 21702-5012 DISTRIBUTION STATEMENT: Approved for Public Release; Distribution Unlimited The views, opinions and/or findings contained in this report are those of the author(s) and should not be construed as an official Department of the Army position, policy or decision unless so designated by other documentation.

REPORT DOCUMENTATION PAGE Form Approved

OMB No. 0704-0188 Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing this collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS. 1. REPORT DATE (DD-MM-YYYY)01-08-2006

2. REPORT TYPEFinal

3. DATES COVERED (From - To)22 Jul 2004 - 21 Jul 2006

4. TITLE AND SUBTITLE MMP-8, a Breast Cancer Bone Metastasis Suppressor Gene

5a. CONTRACT NUMBER

5b. GRANT NUMBER W81XWH-04-1-0687

5c. PROGRAM ELEMENT NUMBER

6. AUTHOR(S) Nagarajan Selvamurugan, Ph.D.

5d. PROJECT NUMBER

5e. TASK NUMBER

E-Mail: selvamn2@umdnj.edu

5f. WORK UNIT NUMBER

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)

8. PERFORMING ORGANIZATION REPORT NUMBER

University of Medicine and Dentistry of New Jersey Piscataway, NJ 08854

9. SPONSORING / MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)U.S. Army Medical Research and Materiel Command

Fort Detrick, Maryland 21702-5012 11. SPONSOR/MONITOR’S REPORT NUMBER(S) 12. DISTRIBUTION / AVAILABILITY STATEMENT Approved for Public Release; Distribution Unlimited

13. SUPPLEMENTARY NOTES

14. ABSTRACT: In order to study the role of MMP-8 on inhibition of cancer growth and progression, we initiated our work to clone the human MMP-8 cDNA and express it in vitro. The MMP-8 cDNA with a V5 epitope tag was cloned downstream into the CMV promoter vector. The construct was verified by sequencing. But the expression level of MMP-8 was not detected by Western blot analysis. The molecular mechanisms of how TGF-β1 mediates stimulation of invasion and formation of bone metastasis have yet to be completely determined. ATF-3 (activating transcription factor-3) was strongly stimulated and its level was sustained by TGF-β1 in highly invasive and bone metastatic breast cancer cells. We have identified for the first time that cyclin A1 and MMP- 13 are ATF-3 target genes. ATF-3 also regulates Runx2 (a bone specific transcription factor) in human breast cancer cells and that may provide a molecular phenotype for ATF-3 to regulate its target genes associated with bone metastasis. 15. SUBJECT TERMS MATRIX METALLOPROTEINASES, TRANSFORMING GROWTH FACTOR, ACTIVATING TRANSCRIPTION FACTOR-3, BONE INVASION AND METASTASIS 16. SECURITY CLASSIFICATION OF:

17. LIMITATION OF ABSTRACT

18. NUMBER OF PAGES

19a. NAME OF RESPONSIBLE PERSONUSAMRMC

a. REPORT U

b. ABSTRACTU

c. THIS PAGEU

UU

33

19b. TELEPHONE NUMBER (include area code)

Standard Form 298 (Rev. 8-98)Prescribed by ANSI Std. Z39.18

Table of Contents

Introduction…………………………………………………………….…………............................3 BODY……………………………………………………………………………………………………4

Key Research Accomplishments………………………………………….……………8 Reportable Outcomes…………………………………………………………………….9 Conclusions………………………………………………………………………………..9 References…………………………………………………………………………………9 Appendices……………………………………………………………………………….11

Selvamurugan, Nagarajan

4

Introduction: TGF-β Signaling TGF-β (transforming growth factor-beta), a multipotent cytokine has a wide range of physiological and pathological effects (1, 2). TGF-β is the most potent known growth inhibitor for epithelial cells (3). Mice with targeted disruption of the Tgfb1 gene develop carcinomas (4). In the breast, loss of TGF-β antiproliferative and apoptotic responses may compromise the turnover of the mammary epithelium, thus favoring tumor formation (5-7). TGF-β signaling involves the type I receptor TβR-I, the type II receptor TβR-II, the regulatory Smads (Smad2 and Smad3), and Smad4 (8). Most of these components have mutations in several human cancers. But, mutations in TGF-β receptors or Smads are rare in breast cancer (9, 10). Moreover, for breast cancer cells, TGF-β1 is a crucial molecule in metastatic breast cancer stimulating invasion (11, 12) and formation of TGF-β-dependent bone metastases in model systems (13). Advances have been made in understanding how TGF-β signals inhibit cell division in normal epithelial cells. The progress of the cell cycle is regulated by the sequential expression of cyclins, followed by the activation of their associated cyclin-dependent kinases (cdks). A number of specific cyclins have been isolated and characterized in mammalian cells, and their temporal patterns of expression have been mapped to specific phases of the cell cycle. ATF-3 Independent observations over the years have defined a small group of immediate TGF-β target genes that contribute to the effect of TGF-β on epithelial cell homeostasis and the importance that its disruption has in cancer (5, 13). Aberrant expression of the AP-2 transcription factor has been linked to the progression of human breast cancer (14). A selective loss of c-myc transcription factor repression by TGF-β1 has been shown in a highly invasive and bone metastatic human breast cancer cell line (MDA-MB231) (15). There is growing evidence indicating that transcription factors such as GADD153, Twist, Runx2, Stat3, NRIF3, TBX3, NF kappaB, DEC1 (16-22) have the ability to alter the progression of breast cancer growth and metastasis and thus, transcription factors are major targets for cancer therapy (23). ATF-3 (activating transcription factor-3), a member of the ATF/CREB subfamily is a bZip transcription factor (24-27). ATF-3 is expressed at very low levels in normal, quiescent cells but can be rapidly and highly induced in different cell types by multiple and diverse extracellular signals (28-30, 31). ATF-3 is a common target of TGF-β1 and stress signals and serves to inhibit cell growth in normal epithelial cells (24). Oligonucleotide microarray analysis showed that TGF-β1 prolonged expression of ATF-3 in breast cancer cells (15). Although the biological functions of ATF-3 have yet to be completely elucidated, there is strong circumstantial evidence that this transcription factor plays an important role in the regulation of normal and neoplastic growth responses. To date, only a few target promoters for ATF-3 (gadd153/CHOP10, cyclin D1 and ATF-3 itself) (30, 32-34) have been identified. The presence of potential ATF-3 binding sites in the promoter regions of other cyclins (35) and of Rb itself (36, 37) suggests that several additional cell cycle-related genes may be subject to regulation by ATF-3. MMP-13 The matrix metalloproteinases (MMPs) are a family of enzymes that are important for tissue remodeling. MMPs, however, also contribute to pathological conditions including rheumatoid arthritis, coronary artery disease, and cancer (38-42). Tumor cells utilize the matrix degrading capability of these enzymes to spread to distant sites. In addition, MMPs promote the growth of these tumor cells once they have metastasized such as lung, brain, and bone. TGF-β1 stimulates MMP-13 (collagenase-3; an invasive and metastasis gene) expression in MDA-MB231 cells and these cells are known to form bone metastases (13, 43, 44). MMP-13 is over-expressed in a variety of malignant tumors (45-49). MMP-13-driven extracellular matrix (ECM) proteolysis may support cancer cell growth both biochemically, by exposing mitogenic factors, and physically, by providing space for the proliferating cells. A greater understanding of the regulatory mechanisms of MMP expression is necessary and will provide several new avenues for therapeutic intervention in controlling breast cancer cell growth, invasion, and metastasis.

Selvamurugan, Nagarajan

5

In order to investigate the role played by ATF-3 in breast cancer cell growth and metastasis, we first examined TGF-β1 regulation of ATF-3 in MCF-10A (normal human mammary epithelial cells) and MDA-MB-231 cells (invasive and bone metastatic human breast cancer cells). TGF-β1 stimulated expression of ATF-3, c-Jun, and JunB in both MCF-10A and MDA-MB231 cells but ATF-3 and c-Jun levels were sustained in MDA-MB231 (Fig. 1). There was no significant change in the level of JunD expression by TGF-β1 after normalization with α-tubulin expression in MCF-10A and MDA-MB231 cells. Body: The specific aim of this proposal was to test if overexpression of MMP-8 in breast cancer cells will contribute to a less aggressive phenotype in breast cancer cells which have metastasized to bone. In order to study the role of MMP-8 on inhibition of cancer growth and progression, we proposed to utilize a transgenic mouse model to overexpress MMP-8 under the control of the bone specific osteocalcin promoter. The osteocalcin promoter has been shown to confer differentiated osteoblast- and post-specific expression to a reporter gene in vivo. To generate transgenic mice overexpressing MMP-8, we first initiated our work to clone the human MMP-8 cDNA (1.4 kbps) and express it in vitro. We used pcDNA3.1 Directional TOP Expression construct (Invitrogen) for this purpose. The pcDNA3.1 contains the following elements: human cytomegalovirus (CMV) immediate-early promoter/enhancer that permits efficient, high-level expression of recombinant protein and V5 epitope that allows detection of recombinant protein with anti-V5 antibody. The MMP-8 cDNA with a V5 epitope tag was cloned downstream into the CMV promoter sequence. The construct was sequenced to verify cloning of the MMP-8 cDNA insert in frame. The construct was transfected into COS-7 cells using the GeneJammer according to the guidelines provided by the company. Cells were lysed and subjected to Western blot analysis. The expression of V5-epitope tagged MMP-8 protein was not detected by the V5-epitope antibody. One of our collaborators (Dr. Susan Rittling) generated a series of metastatic murine mammary epithelial cell lines using normal mice rather than using nude mice. Cardiac injection of mouse mammary pad tumor cell line r3T into 129 strain female mice leads to development of bone metastases (50). To get expertise in the techniques of cardiac injection of cancer cells into the mice and tumor analysis, we utilized those cancer cells with normal mice. Fig 1. Osteolytic bone metastases. Female mice were sacrificed three weeks after left ventricle injection of medium alone or medium containing breast cancer cells (r3T). Bones were dissected and cleaned of soft tissues, and visualized by X-ray. Arrows indicate regions of bone loss. Fig 2. Histological appearance of metastatic tumor cells. Bones were decalcified in EDTA, embedded in paraffin and stained with Gomori trichrome. (A) distal femur, control mouse. (B) distal femur, mouse with osteolytic metastasis. Note replacement of entire marrow with tumor cells. TGF-β1 is the most potent known growth inhibitor for epithelial cells. In breast tissue, loss of TGF-β1 anti-proliferative response favors tumor formation. Moreover, in breast cancer cells, TGF-β1 is a crucial molecule for stimulation of invasion and formation of bone metastases. The molecular mechanisms of how TGF-β1 mediates these effects have yet to be completely determined. In my laboratory, we have found a defect in repression of ATF-3 (activating transcription factor-3) expression by TGF-β1 in bone metastatic human breast cancer (MDA-MB231) and mouse mammary pad tumor (r3T) cells.

Selvamurugan, Nagarajan

6

Fig 3. TGF-β1 stimulates expression of ATF-3, c-Jun, and JunB. MCF-10A (normal human mammary cells) and MDA-MB231 cells were treated with control or TGF-β1 (1 ng/ml)-containing media for the indicated times. Total lysates were prepared and subjected to Western blot analysis using the antibodies shown in the figure. α-tubulin represents the loading control.

30’ 1 h 2 h 4 h

MCF-10A

a26-

37-

37-

50-

kD

MDA-MB231

ATF3

c-Jun

JunB

α-tubulin

JunD37-

C T C T C T C T30’ 1 h 2 h 4 h

C T C T C T C T

TGF-β1 stimulated expression of ATF-3 and its level was sustained even at 24 h in r3T cells (Fig. 4). In contrast, in normal murine mammary glandular epithelial cells (NMuMG), ATF-3 expression peaked at 2 h after TGF-β1 and then declined (data not shown). In normal mammary epithelial cells, TGF-β1 stimulates a transient expression of ATF-3; whereas in metastatic mammary epithelial cells, TGF-β1 stimulates a sustained expression of ATF-3. The sustained and prolonged expression of ATF-3 may lead to alterations in homo- and heterodimerization of AP-1 family members and other proteins, which could activate the genes that participate in breast cancer progression. Fig 4. TGF-β1 stimulates ATF-3 expression. Mouse mammary pad tumor cells (r3T) were treated with control or TGF-β1 (1 ng/ml)-containing media for the indicated times. Total lysates were prepared and subjected to Western blot analysis using the antibodies shown in the figure. α-tubulin represents the loading control.

C T C T C T C T C T C T 30’ 1 h 2 h 4 h 8 h 24 h

kDa

20 -

50 -

ATF-3

α-tubulin

We wanted to detect the level of ATF-3 protein in breast tissues obtained from normal and cancer patients. Tissue microarray slides containing normal human breast tissues and human primary breast cancer tissues (Imgenex, CA) were processed for immunohistochemical staining (a kit from LabVision, CA) with ATF-3 antibody. The results (Fig. 5) clearly indicated that ATF-3 expression is very low in normal breast tissue and is high in breast cancer tissues.

(A) (B) (C)Fig 5. ATF-3 expression in human primary breast carcinomas. Histospot staining in tissue microarrays include low level staining in normal breast tissue (A) and strong staining in breast cancer tissues (B and C) of two different patients (40X magnification). ATF-3 protein is represented by brown color staining. Since ATF-3 is highly expressed in human breast cancer cells (Fig. 1) and mouse mammary pad tumor cells (Fig. 2) and human primary tumors (Fig. 3), we wanted to first determine whether overexpression of ATF-3 is sufficient to induce cellular proliferation in vitro. We transiently transfected the ATF-3 eukaryotic expression plasmid (pCMV-ATF-3) into normal human mammary epithelial cells (MCF-10A). The empty eukaryotic expression plasmid (pCMV) was also transfected as a control for transfection effects. The cells were counted using a haemocytometer after 6 days (~80% confluency). In parallel, cells were also pulsed with 0.5 µCi/ml [3H]thymidine for 4 h before harvesting and assessing radioactivity. The results indicated that overexpression of ATF-3 increases normal human mammary epithelial cell number (Fig. 6A) and DNA synthesis (Fig. 6B) over control (empty vector).

Selvamurugan, Nagarajan

7

C T C T

scram siRNA ATF-3 siRNA

50 -

M(kDa)

26 - ATF-3

α-tubulin

Fig 6. ATF-3 increases normal human mammary epithelial cell growth. (A) Cells were trypsinized, harvested, and counted by haemocytometer. (B) Cell proliferation in terms of DNA synthesis was assessed by measuring the incorporation of 3H-thymidine into cells for a 4 h period. Cells were harvested onto glass fiber filters using an automated cell harvester, and counted in a Packard liquid scintillation counter. Data represent mean ± S.E. of three experiments. To determine the functional role of ATF-3 in breast cancer progression, we used the RNA interference technique for in vivo depletion of a gene product. The hairpin oligonucleotides that target ATF-3 (ATF-3 siRNA) or nonspecific sequences (scrambled siRNA) were cloned into the psiSTRIKE U6 hairpin vector (Promega). Transient transfection of MDA-MB231 cells with the psiSTRIKE vector that contained hairpin oligonucleotides with a human ATF-3 target sequence decreased both the basal and TGF-β1-stimulated ATF-3 expression, compared with the nonspecific target sequences (Fig. 7). Fig 7. ATF-3 siRNA reduces both basal and TGF-β1-stimulated ATF-3 expression. MDA-MB231 cells were transiently transfected with either scrambled siRNA or ATF-3 siRNA vectors for 24 h and then treated with control or TGF-β1 (1 ng/ml)-containing media for 4 h. Total lysates were prepared and subjected to Western blot analysis using the ATF-3 and α-tubulin (loading control) antibodies. Since TGF-β loses its antiproliferative activity in breast cancer cells, we determined whether knockdown of ATF-3 expression has any effect on expression of the cell cycle genes. As shown in Fig. 8, TGF-β1 stimulated expression of cyclin A1, -B1, -D1, and -E in these cells while ATF-3 siRNA only decreased expression of cyclin A1 in both control and TGF-β1-stimulated MDA-MB231 cells. Thus, ATF-3 must be the mediator of TGF-β1-stimulation of cyclin A1 and cyclin A1 is likely to be an ATF-3 target gene. Cyclin A1, an alternative A-type cyclin that is essential for spermatogenesis contributes cell cycle progression from G1 to S phase (51) and is also expressed in hematopoietic progenitor cells and in acute myeloid leukemia.

cyclin A1

C T C Tscram siRNA ATF-3 siRNA

50 -

M(kDa)

Fig 8. ATF-3 siRNA reduces both basal and TGF-β1-stimulated cyclin A1 expression. MDA-MB231 cells were transiently transfected with either scrambled siRNA or ATF-3 siRNA plasmids for 24 h and then treated with control or TGF-β1 (1 ng/ml)-containing media for 24 h. Total lysates were prepared and subjected to Western blot analysis using the antibodies as indicated. α-tubulin was a loading control.

cyclin B1

cyclin D1

cyclin E

α-tubulin50 -

35 -

50 -

50 - Since MDA-MB231 cells are highly invasive and bone metastatic in nature and TGF-β1 stimulates MMP-13 (an invasive and metastatic gene) (43) and ATF-3 expression in these cells, we next determined whether MMP-13 is a target gene for ATF-3. The -148 MMP-13 promoter that contains 148 base pairs upstream of the transcription initiation site retains the TGF-β-responsive region (44). The -148 MMP-13 promoter fused with a reporter gene, chloramphenicol acetyl transferase (CAT) was transiently transfected with either scrambled siRNA or ATF siRNA constructs into MDA-MB231 cells. As shown in Fig. 9, TGF-β1 stimulated MMP-13 promoter activity and ATF-3 siRNA reduced both the control and TGF-β1-stimulated MMP-13 promoter activity in these cells. Hence, the MMP-13 gene (another potential ATF-3 target gene) is regulated by TGF-β1 via ATF-3.

Selvamurugan, Nagarajan

8

Fig 9. ATF-3 siRNA reduces both the basal and TGF-β1-stimulated MMP-13 promoter activity. The wild type MMP-13 promoter construct (-148) was transiently cotransfected with either srambled siRNA or ATF-3 siRNA plasmids into MDA-MB231 cells for 24 h and then treated with control or TGF-β1 (1 ng/ml)-containing media for 24 h. Lysates were prepared and assayed for CAT activity. Renella luciferase was used to normalize the transfection efficiency. Data represent mean ± S.E. of three experiments. MDA-MB231 cells are highly bone metastatic in nature. Runx2, a bone specific transcription factor responsible for expression of bone marker genes including MMP-13 is regulated by TGF-β1 in these cells (44). Hence, we next determined whether Runx2 could be a target gene for ATF-3. MDA-MB231 cells were transiently transfected with either scrambled siRNA or ATF-3 siRNA vectors and lysates were prepared and subjected to Western blot analysis. As shown in Fig. 10, ATF-3 siRNA decreased expression of Runx2 proteins (65 kDa and 50 kDa) in human breast cancer cells, indicating that Runx2 may act as a direct ATF-3 target gene. Fig 10. ATF-3 regulation of Runx2. MDA-MB231 cells were transiently transfected with either scrambled siRNA or ATF-3 siRNA plasmids for 24 h. Total lysates were prepared and subjected to Western blot analysis using the Runx2 antibody. Key Research Accomplishments:

• The construct pCMV-V5-MMP-8 was made. • Cardiac injection of tumor cells into mice and histology of bone metastasized cells were standardized. • TGF-β1 stimulated prolonged and sustained expression of ATF-3 protein in the human breast cancer cell

line MDA-MB231 and in the mouse mammary gland cancer cell line r3T. • TGF-β1 stimulated expression of cyclin A1, B1, D1 and E in MDA-MB231 cells while ATF-3 siRNA

decreased only expression of cyclin A1 in both control and TGF-β1-stimulated MDA-MB231 cells. • ATF-3 must be the mediator of TGF-β1-stimulation of cyclin A1 and Cyclin A1 is likely to be an ATF-3

target gene. • TGF-β1 stimulated MMP-13 promoter activity and ATF-3 siRNA reduced both the control and TGF-β1-

stimulated MMP-13 promoter activity in MDA-MB231 cells. • MMP-13 is another potential ATF-3 target gene and regulated by TGF-β1 via ATF-3. • ATF-3 siRNA decreased expression of Runx2 proteins, suggesting that Runx2 is another target gene for

ATF-3. Reportable Outcomes: Manuscript: 1. Parathyroid hormone stimulation and PKA signaling of latent transforming growth factor-beta binding protein-1 (LTBP-1) mRNA expression in osteoblastic cells. S. Kwok, L. Qin, N. C. Partridge, and N. Selvamurugan (2005) * corresponding author Journal of Cellular Biochemistry 95:1002-11

Selvamurugan, Nagarajan

9

2. Overexpression of Runx2 directed by the matrix metalloproteinase-13 promoter containing the AP-1 and Runx/RD/Cbfa sites alters bone remodeling in vivo. N. Selvamurugan, Jefcoat, S. Jr., S. Kwok, Y. Yang, R. Kowalewski, J. Tamasi, and N. C. Partridge (2006) Journal of Cellular Biochemistry, April 25 (Epub ahead of print)

Abstract: TGF-β1 Regulation of ATF-3 and its Target Genes in Bone Metastasizing Breast Cancer Cells. Presented at the 27th Annual meeting of American Society for Bone and Mineral Research, Nashville, TN, on September 23-27, 2005. Conclusions:

1. The application of a transgenic mouse model will contribute greatly to the understanding of the pathogenesis of bone metastasis. Identification of the exact nature of these tumor-bone interactions may not only generate valuable information on underlysing regulatory mechanisms in invasion and bone metastasis but can also be of value in the development of therapeutic strategies.

2. We are the first to identify the TGF-β1-regulation of ATF-3 and its target genes, cyclin A1, MMP-13,

and Runx2 in bone metastasizing breast cancer cells. The dysregulation of ATF-3 by TGF-β1 in breast cancer cells may be key to the subsequent metastasis of these cells to bone.

References: 1 Massague, J. and Wotton, D. (2000) EMBO Journal 19, 1745-54 2 Piek, E., Heldin, C. H. and Ten Dijke, P. (1999) FASEB Journal 13, 2105-24 3 Piek, E. and Roberts, A. B. (2001) Advances in Cancer Research 83, 1-54 4 Engle, S. J., Hoying, J. B., Boivin, G. P., Ormsby, I., Gartside, P. S. and Doetschman, T. (1999) Cancer

Research 59, 3379-86 5 Massague, J., Blain, S. W. and Lo, R. S. (2000) Cell 103, 295-309 6 Barcellos-Hoff, M. H. and Ewan, K. B. (2000) Breast Cancer Research 2, 92-9 7 Nguyen, A. V. and Pollard, J. W. (2000) Development 127, 3107-18 8 Massague, J. (2000) Nature Reviews Molecular Cell Biology 1, 169-78 9 Tomita, S., Deguchi, S., Miyaguni, T., Muto, Y., Tamamoto, T. and Toda, T. (1999) Breast Cancer

Research & Treatment 53, 33-9 10 Anbazhagan, R., Bornman, D. M., Johnston, J. C., Westra, W. H. and Gabrielson, E. (1999) Cancer

Research 59, 3363-4 11 Welch, D. R., Fabra, A. and Nakajima, M. (1990) Proceedings of the National Academy of Sciences of

the United States of America 87, 7678-82 12 Oft, M., Peli, J., Rudaz, C., Schwarz, H., Beug, H. and Reichmann, E. (1996) Genes & Development 10,

2462-77 13 Yin, J. J., Selander, K., Chirgwin, J. M., Dallas, M., Grubbs, B. G., Wieser, R., Massague, J., Mundy, G.

R. and Guise, T. A. (1999) Journal of Clinical Investigation 103, 197-206 14 Turner, B. C., Zhang, J., Gumbs, A. A., Maher, M. G., Kaplan, L., Carter, D., Glazer, P. M., Hurst, H.

C., Haffty, B. G. and Williams, T. (1998) Cancer Research 58, 5466-72 15 Chen, C. R., Kang, Y. and Massague, J. (2001) Proceedings of the National Academy of Sciences of the

United States of America 98, 992-9 16 Talukder, A. H., Wang, R. A., and Kumar, R. (2002) Oncogene. 21, 4289-4300 17 Yang, J., Mani, S. A., Donaher, J. L., Ramaswamy, S., Itzykson, R. A., Come, C., Savagner, P.,

Gitelman, I., Richardson, A., and Weinberg, R. A. (2004) Cell. 177, 927-939 18 Barnes, G. L., Hebert, K. E., Kamal, M., Javed, A., Einhorn, T. A., Lian, J. B., Stein, G. S., and

Gerstenfeld, L. C. (2004) Cancer Res. 64, 4506-4513

Selvamurugan, Nagarajan

10

19 Dechow, T. N., Pedranzini, L., Leitch, A., Leslie, K., Gerald, W. L., Linkov, I., and Bromberg, J. F. (2004) Proceedings of the National Academy of Sciences of the United States of America 101, 10602-10607

20 Fan, W., Huang, X., Chen, C., Gray, J., and Huang, T. (2004) Cancer Research 64, 5132-5139 21 Huber, M. A., Azoitei, N., Baumann, B., Grunert, S., Sommer, A., Pehamberger, H., Kraut, N., Beug,

H., and Wirth, T. (2004) J Clin Invest. 114, 569-581 22 Chakrabarti, J., Turley, H., Campo, L., Han, C., Harris, A. L., Gatter, K. C., and Fox, S. B. (2004) Br J

Cancer 91, 954-958 23 Darnell, J. E. (2002) Nature Medicine 8, 1068-1071 24 Kang, Y., Chen, C. R. and Massague, J. (2003) Molecular Cell 11, 915-26 25 Chen, B. P., Liang, G., Whelan, J. and Hai, T. (1994) Journal of Biological Chemistry 269, 15819-26 26 Hsu, J. C., Bravo, R. and Taub, R. (1992) Molecular & Cellular Biology 12, 4654-65 27 Hsu, J. C., Cressman, D. E. and Taub, R. (1993) Cancer Research 53, 3789-94 28 Weir, E., Chen, Q., DeFrances, M. C., Bell, A., Taub, R. and Zarnegar, R. (1994) Hepatology 20, 955-

60 29 Drysdale, B. E., Howard, D. L. and Johnson, R. J. (1996) Molecular Immunology 33, 989-98 30 Hai, T., Wolfgang, C. D., Marsee, D. K., Allen, A. E. and Sivaprasad, U. (1999) Gene Expression 7,

321-35 31 Xie, J., Bliss, S. P., Nett, T. M., Ebersole, B. J., Sealfon, S. C. and Roberson, M. S. (2005) Molecular

Endocrinology [Jun 16; Electronic publication ahead of print) 32 Wolfgang, C. D., Chen, B. P., Martindale, J. L., Holbrook, N. J. and Hai, T. (1997) Molecular &

Cellular Biology 17, 6700-7 33 Wolfgang, C. D., Liang, G., Okamoto, Y., Allen, A. E. and Hai, T. (2000) Journal of Biological

Chemistry 275, 16865-70 34 Allan, A. L., Albanese, C., Pestell, R. G. and LaMarre, J. (2001) Journal of Biological Chemistry 276,

27272-80 35 Yoshizumi, M., Hsieh, C. M., Zhou, F., Tsai, J. C., Patterson, C., Perrella, M. A. and Lee, M. E. (1995)

Molecular & Cellular Biology 15(6):3266-72 36 Sakai, T., Ohtani, N., McGee, T. L., Robbins, P. D. and Dryja, T. P. (1991) Nature 353, 83-6 37 Linardopoulos, S., Gonos, E. S. and Spandidos, D. A. (1993) Cancer Letters 71, 67-74 38 Birkedal-Hansen, H., Moore, W. G., Bodden, M. K., Windsor, L. J., Birkedal-Hansen, B., DeCarlo, A.

and Engler, J. A. (1993) Critical Reviews in Oral Biology & Medicine 4, 197-250 39 Stetler-Stevenson, W. G., Liotta, L. A. and Kleiner, D. E., Jr. (1993) FASEB Journal 7, 1434-41 40 Crawford, H. C. and Matrisian, L. M. (1994) Invasion & Metastasis 14, 234-45 41 Crawford, H. C. and Matrisian, L. M. (1996) Enzyme & Protein 49, 20-37 42 Benaud, C., Dickson, R. B. and Thompson, E. W. (1998) Breast Cancer Research & Treatment 50, 97-

116 43 Selvamurugan, N., Fung, Z. and Partridge, N. C. (2002) FEBS Letters 532, 31-5 44 Selvamurugan, N., Kwok, S. and Partridge, N. C. (2004) Journal of Biological Chemistry 279, 27764-73 45 Barrett, J. M., Puglia, M. A., Singh, G. and Tozer, R. G. (2002) Breast Cancer Research & Treatment

72, 227-32 46 Freije, J. M., Balbin, M., Pendas, A. M., Sanchez, L. M., Puente, X. S. and Lopez-Otin, C. (2003)

Advances in Experimental Medicine & Biology 532, 91-107 47 Hattori, Y., Nerusu, K. C., Bhagavathula, N., Brennan, M., Hattori, N., Murphy, H. S., Su, L. D., Wang,

T. S., Johnson, T. M. and Varani, J. (2003) Experimental & Molecular Pathology 74, 230-7 48 Ohshiba, T., Miyaura, C., Inada, M. and Ito, A. (2003) Br J Cancer. 88, 1318-26 49 Pang, S. T., Flores-Morales, A., Skoog, L., Chuan, Y. C., Nordstedt, G. and Pousette, A. (2004) 50 Chen, Y. and Rittling, S. R. (2003) Clinical and Experimental Metastasis 20, 111-20 51 Ji, P., Agrawal, S., Diederichs, S., Baumer, N., Becker, A., Cauvet, T., Kowski, S., Beger, C., Welte, K.,

Berdel, W. E., Serve, H. and Muller-Tidow, C. (2005) Oncogene 24, 2739-44

Selvamurugan, Nagarajan

11

Appendices: 1. Parathyroid hormone stimulation and PKA signaling of latent transforming growth factor-beta binding protein-1 (LTBP-1) mRNA expression in osteoblastic cells. S. Kwok, L. Qin, N. C. Partridge, and N. Selvamurugan (2005) * corresponding author Journal of Cellular Biochemistry 95:1002-11 2. Overexpression of Runx2 directed by the matrix metalloproteinase-13 promoter containing the AP-1 and Runx/RD/Cbfa sites alters bone remodeling in vivo. N. Selvamurugan, Jefcoat, S. Jr., S. Kwok, Y. Yang, R. Kowalewski, J. Tamasi, and N. C. Partridge (2006) Journal of Cellular Biochemistry, April 25 (Epub ahead of print)

Journal of Cellular Biochemistry 95:1002–1011 (2005)

Parathyroid Hormone Stimulation and PKA Signaling ofLatent Transforming Growth Factor-b Binding Protein-1(LTBP-1) mRNA Expression in Osteoblastic Cells

Sukyee Kwok, Ling Qin, Nicola C. Partridge, and Nagarajan Selvamurugan*

Department of Physiology and Biophysics, University of Medicine and Dentistry of New Jersey,Piscataway, New Jersey 08854

Abstract Parathyroidhormone (PTH) regulates bone remodeling andcalciumhomeostasis by acting onosteoblasts.Recently, the gene expression profile changes in the rat PTH (1–34, 10�8M)-treated rat osteoblastic osteosarcoma cellline, UMR 106-01, using DNA microarray analysis showed that mRNA for LTBP-1, a latent transforming growth factor(TGF-b)-binding protein is stimulated by PTH. Latent TGF-b binding proteins (LTBPs) are required for the proper foldingand secretion of TGF-b, thus modifying the activity of TGF-b, which is a local factor necessary for bone remodeling. Weshowhere by real timeRT-PCR that PTH-stimulated LTBP-1mRNAexpression in rat andmouse preosteoblastic cells. PTHalso stimulated LTBP-1mRNA expression in all stages of rat primary osteoblastic cells but extended expression was foundin differentiating osteoblasts. PTH also stimulated TGF-b1mRNA expression in rat primary osteoblastic cells, indicating alink between systemic and local factors for intracellular signaling in osteoblasts. An additive effect on LTBP-1 mRNAexpression was found when UMR 106-01 cells were treated with PTH and TGF-b1 together. We further examined thesignaling pathways responsible for PTH-stimulated LTBP-1 and TGF-b1mRNA expression inUMR 106-01 cells. The PTHstimulation of LTBP-1 and TGF-b1 mRNA expression was dependent on the PKA and the MAPK (MEK and p38 MAPK)pathways, respectively in these cells, suggesting that PTH mediates its effects on osteoblasts by several intracellularsignaling pathways. Overall, we demonstrate here that PTH stimulates LTBP-1mRNA expression in osteoblastic cells andthis is PKA-dependent. This event may be important for PTH action via TGF-b in bone remodeling. J. Cell. Biochem. 95:1002–1011, 2005. � 2005 Wiley-Liss, Inc.

Key words: PTH; TGF-b; LTBP-1; osteoblast; PKA signaling

Parathyroid hormone (PTH) is one of themajor calciotropic hormones affecting serumcalcium levels and bone remodeling [Tam et al.,1982; Dobnig and Turner, 1997; Swarthoutet al., 2002]. PTH acts by binding to the PTH1R,a G-protein-coupled receptor on osteoblasts,resulting in functional changes in the actionsof both osteoblasts [Bellows et al., 1990] andosteoclasts [Kanzawa et al., 2000; Qin et al.,2004]. The molecular mechanisms regulatingthe activities of both cells by PTH are still notcompletely known.

There is growing evidence that growth factorsand cytokines released from bone matrix playimportant roles in the coupling of bone resorp-tion to bone formation and in repair processessuch as fracture healing. Bone extracellularmatrix (ECM) is the major storage site in thebody for transforming growth factor-beta (TGF-b), which is a multipotent cytokine [Seyedinet al., 1985; Hauschka et al., 1986]. TGF-b is

� 2005 Wiley-Liss, Inc.

Abbreviations used: PTH, parathyroid hormone; LTBP-1,latent transforming growth factor-beta binding protein-1;TGF-b, transforming growth factor-beta; ECM, extracellularmatrix; SLC, small latent TGF-b complex; LLC, large latentTGF-b complex; LAP, latency associated protein; FBS, fetalbovine serum; RT-PCR; reverse transcriptase polymerasechain reaction; MAPK, mitogen-activated protein kinase;PKA, protein kinase A; PKC, protein kinase C.

Grant sponsor: Department of Defense; Grant number:W81XWH-04-1-0687; Grant sponsor: New Jersey Commis-sion on Cancer Research; Grant sponsor: Foundation of theUniversity of Medicine and Dentistry of New Jersey; Grantsponsor: NIH; Grant numbers: DK48109, DK47420.

*Correspondence to: Nagarajan Selvamurugan, PhD,Department of Physiology and Biophysics, UMDNJ-RobertWood Johnson Medical School, 675 Hoes Lane, Piscataway,NJ 08854. E-mail: selvamn2@umdnj.edu

Received 24 November 2004; Accepted 28 January 2005

DOI 10.1002/jcb.20453

synthesized as a homodimeric pro-protein andthe dimeric pro-peptide is cleaved intracellu-larly from the growth factor [Taipale et al.,1994]. The TGF-b propeptide binds to TGF-b,and the proteins are secreted as a complex[Annes et al., 2003]. In this small latentcomplex (SLC), TGF-b cannot bind to its surfacereceptors. Therefore, the propeptide is calledthe latencyassociatedprotein (LAP).TheSLC issecreted by bone cells [Bonewald et al., 1991],chondrocytes [Pedrozo et al., 1999], kidney cells[Marra et al., 1996], and prostate cells [Dallaset al., 2005]. The dissociation or activation ofTGF-b fromLAP is a critical regulatory event asall TGF-b is secreted in a latent form. The LAPdimer is usually disulfide bonded to a secondgene product, latent TGF-b binding protein(LTBP), and the trimolecular aggregate is calledthe large latent complex (LLC) [Rifkin, 2005].A major mechanism for storage of secreted

latent TGF-b in bone matrix is via its associ-ation with the latent TGF-b binding protein-1(LTBP-1) [Taipale et al., 1994;Dallas etal., 1995;Dallas et al., 2000]. LTBP is a member of theLTBP/fibrillin protein family, which comprisesfibrillin-1, fibrillin-2 and fibrillin-3, and LTBP-1,LTBP-2, LTBP-3, and LTBP-4 [Ramirez andPereira, 1999; Oklu and Hesketh, 2000]. LTBPsare required for the proper folding and secretionof TGF-b, thus modifying the activity of TGF-b[Miyazono et al., 1991]. In human and rats,LTBP-1 appears as two mRNA species, whichencode for two different NH2-terminal variants,the longer LTBP-1L having a 346 amino acidextension not present in the shorter LTBP-1Sisoform [Kanzaki et al., 1990; Saharinen et al.,1999]. Both isoforms possess their own, indepen-dent promoter regions, capable of regulating thetissue type specific expression of LTBP-1 iso-forms [Koski et al., 1999]. The LLC-containingthe LTBP-1L is found in ECM [Kanzaki et al.,1990];whereas theLLC-containing theLTBP-1Sis found in platelets [Wakefield et al., 1988]. ThisECM-boundTGF-b stored in a latent formcan bereleased and activated by resorbing osteoclasts[Oreffo et al., 1989; Oursler, 1994]. Oncereleased from the matrix and activated, TGF-bcan influence inhibition of osteoclast activity,osteoblast proliferation, and stimulation of pro-duction of bone ECM proteins [Hughes et al.,1996; Roberts, 1998; Bonewald, 1999]. TGF-bhas therefore been implicated as a couplingfactor that coordinates the processes of boneresorption and subsequent bone formation.

The rat osteoblastic cell line, UMR 106-01 is auseful model system for studying the effects ofPTH on osteoblastic cells in vitro. Recently, thegene expression profile changes in these cellstreated with rat PTH (1–34, 10�8M) usingDNAmicroarray analysis have been published [Qinet al., 2003]. LTBP-1 expressionwas stimulatedby PTH in these cells. Since PTH stimulatesLTBP-1mRNAexpression and that controls theactivity of TGF-b, LTBP-1 seems to be a media-tor in controlling PTH action on osteoblasts viaTGF-b. In this study, we show PTH stimulationof LTBP-1 mRNA expression in the mouse andrat osteoblastic cell lines and in proliferating,differentiating, and mineralizing rat primaryosteoblasts. We have also identified the signal-ing pathways used by PTH in stimulation ofLTBP-1 mRNA and TGF-b1 mRNA expressionin rat osteoblastic cells.

MATERIALS AND METHODS

Materials

Rat PTH (1–34) and human TGF-b1 werepurchased from Sigma, St. Louis, MO andPromega, Madison, WI, respectively. Syntheticoligonucleotides were synthesized by Invitro-gen, Carlsbad, CA. Tissue culture medium andreagents were also obtained from Invitrogen.The MEK1/2, p38 MAPK, JNK, PKA, and PKCinhibitors were purchased from Calbiochem,San Diego, CA. All other chemicals were ob-tained from Sigma.

METHODS

Cell Culture

The rat osteoblastic cells (UMR 106–01) andthe mouse preosteoblastic cells (MC3T3) weremaintained in monolayer in Eagle’s minimalessential medium (with Earle’s salts; EMEM)supplemented with nonessential amino acids,25 mMHEPES (pH 7.3), 5% fetal bovine serum(FBS), 100 units/ml penicillin, and 100 mg/mlstreptomycin at 378C in a humidified atmo-sphere of 5% CO2 and 95% air.

Rat Primary Osteoblastic Cells

Rat primary osteoblasts were isolated by themethod of Shalhoub et al. [1992]. Osteoblastswere derived from postnatal day 1 rat calvariaeby sequential digestions of 20, 40, and 90min at378C in 2 mg/ml collagenase A, 0.25% trypsin.Cells from digests one and two were discarded.

LTBP-1, a Mediator for PTH Action Via TGF-b on Osteoblasts 1003

Cells from the third digest were plated at6.4� 103 cells/cm2 and grown inminimal essen-tial medium (MEM) supplemented with 10%FBS. After reaching confluence (day 7), themedium was switched to BGJb with 10% FBScontaining 50 mg/ml ascorbic acid and 10 mMb-glycerophosphate to allow for initiation ofdifferentiation and mineralization. Mediumchanges were performed every 2 days. Thedetermination of proliferating, differentiating,and mineralizing stages of osteoblasts has beenestablished by demonstration of alkaline phos-phatase activity, osteocalcin production, ali-zarin red staining, and a sensitive adenylylcyclase response to PTH [Shalhoub et al., 1992;Winchester et al., 1999].

Total RNA Isolation and Real Time ReverseTranscriptase-PCR

The rat osteoblastic cells and the mousepreosteoblastic cells were treated with eitherrat PTH (1–34, 10�8M) or human TGF-b1 (1 ng/ml) or both together for different time periods.To determine de novo protein synthesis, cellswere pretreated with cycloheximide (30 mg/ml)for 1 h before PTH treatment. To determine thesignaling pathways, cells were pretreated withDMSO, PD98059, SB203580, SP600125, H89,or GF109203X for 20 min before PTH treat-ment. Cells were rinsed once with 10 ml of cold(48C) PBS, pH 7.4, and harvested. Total RNAwas isolated using the QIAGEN RNeasy Minikit. Reverse transcription was carried out usingTaqMan reverse transcription reagents (RocheApplied Science, Indianapolis, IN). PCRs wereperformed using a real time PCR DNA OpticonEngine (MJ Research, Inc., Watertown, MA)according to the manufacturer’s instructions,which allow real time quantitative detection ofthe PCR product by measuring the increase inSYBR green fluorescence caused by binding ofSYBR green to double-stranded DNA. Eachanalysis was performed three to four times withindependent sets of cells. The data are repre-sented asmean�SEM. Statistical analysis wasperformed by Student’s t-test. Primers for ratLTBP-1, TGF-b1, MMP-13, and b-actin weredesigned using Primer Express software(PerkinElmer Life Sciences). The sequences ofthe above primers were as follows: LTBP-1:Forward, 50-CGTGGCTGGAATGGACAATG,Re-verse, 50-TGGTCTGGTGTGGGGCTGTA;TGF-b1: Forward, 50-TTAGGAAGGACCTGGGTTG-GA, Reverse, 50-ACTGTGTGTCCAGGCTCCA-

AAT;MMP-13: Forward, 50-GCCCTATCCCTT-GATGCCATT, Reverse, 50-ACAGTTCAGGCT-CAACCTG; b-actin: Forward, 50-TCCTGAGC-GCAAGTACTCTGTG, Reverse, 50-CGGACTC-ATCGTACTCCTGCTT.

RESULTS

PTH Stimulates LTBP-1 mRNA Expression

To study the effect of PTH on expression ofLTBP-1 in the rat osteoblastic osteosarcomaline UMR 106-01, cells were treated with ratPTH (1–34) either for different time periodswith 10�8M concentration (Fig. 1A) or at dif-ferent concentrations for 4 and 24 h (Fig. 1B).Total cellular RNAs were purified and analyzedby real time RT-PCR using specific primers forratLTBP-1andb-actins.As shown inFigure1A,LTBP-1 mRNA expression was maximallystimulated (20-fold) by 10�8M PTH at 4 h inUMR 106-01 cells and was still significant at24 h. A wide range of PTH concentrations from10�9 to 10�7M stimulated LTBP-1 mRNA ex-pression at 4 h in these cells but with 10�7 and10�8M PTH concentration, the fold stimulationof LTBP-1 mRNA expression was maintainedout to 24 h (Fig. 1B).

PTH Stimulates LTBP-1 mRNA Expression inMouse Preosteoblastic and Rat Primary

Osteoblastic Cells

Wenext determinedPTH-stimulatedLTBP-1mRNA expression in other osteoblastic cells.The MC3T3 mouse preosteoblastic cells weretreated with control or rat PTH (1–34, 10�8M)-containing media for 1, 4, 12, and 24 h. TotalRNA was isolated and examined for LTBP-1mRNA expression by real time RT-PCR analy-sis. PTH significantly stimulated LTBP-1mRNA expression at 4 h in these cells (Fig. 2A)but the fold stimulation was less than that seenin UMR 106-01 cells (Fig. 1A). To determineexpression of PTH-regulated LTBP-1 in ratprimary osteoblastic cells, proliferating, differ-entiating, and mineralizing cells were treatedwith either control or rat PTH (1–34, 10�8M)-containing media for different times. TotalRNAwas isolated and subjected to examinationfor LTBP-1 mRNA expression by real timeRT-PCR analysis. Our results (Fig. 2B) showthat LTBP-1 mRNA expression was sti-mulated by PTH in proliferating (1 h) andmineralizing (4h) osteoblasts. Indifferentiatingosteoblasts, LTBP-1 mRNA expression was

1004 Kwok et al.

found at 1 h and its level persisted up to 4 hwithPTH-treatment.

PTH Stimulation of LTBP-1 mRNA ExpressionIs a Primary Effect

To determine whether the PTH-mediatedincrease in LTBP-1 mRNA is a primary res-ponse, UMR 106-01 cells were treated withcontrol medium or medium containing rat PTH(1–34, 10�8M) for 4 h in the presence or absenceof 30 mg/ml cycloheximide added 1 h beforetreatment.TotalRNAwas subjected to real timeRT-PCR analysis using specific primers for ratLTBP-1 and b-actin. As shown in Figure 3A,cycloheximide did not inhibit PTH induction ofLTBP-1 mRNA expression, indicating that thePTH stimulation of LTBP-1 expression is aprimary effect and de novo protein synthesis isnot required for this purpose. In fact, cyclohex-

imide increased the PTH-response, indicatingthe inhibition of de novo synthesis of repressorproteins for this effect. As a positive control forcycloheximide treatment, we analyzed mRNAexpression of MMP-13 (matrix metalloprotei-nase-13; collagenase-3) in UMR 106-01 cells(Fig. 3B). We have previously shown that thePTH stimulation of MMP-13 expression is asecondary effect in these cells [Scott et al.,1992].

PTH Stimulates TGF-b1 mRNA Expression andBoth PTH and TGF-b1 Have an Additive

Effect on LTBP-1 mRNA Expression

Since LTBPs are required for the properfolding and secretion of TGF-b [Miyazono et al.,1991], the increased LTBP-1 expression causedby PTH may be correlated with increasedexpression of TGF-b. Hence, we determined if

Fig. 1. Effect of PTH on expression of LTBP-1 mRNA levels inthe rat osteoblastic cell line, UMR 106-01. A: Time course of thePTH stimulation of LTBP-1. UMR 106-01 cells were serum-starved for 24 h and treated with control medium or mediumcontaining rat PTH (1–34, 10�8M) for different times asindicated. Total RNA was isolated and subjected to real timeRT-PCR using specific primers for rat LTBP-1 and b-actin. B:Concentration-dependence of the PTH stimulation of LTBP-1

mRNA. UMR 106-01 cells were serum-starved for 24 h andtreated with control medium or medium containing PTH atdifferent concentrations as indicated for 4 or 24 h. Total RNAwasisolated and subjected to real time RT-PCRusing specific primersfor rat LTBP-1 and b-actin. The relative levels of mRNAs werenormalized to b-actin, and the PTH-fold changes were calcu-lated over controls. The asterisks represent P<0.05 comparedwith control.

LTBP-1, a Mediator for PTH Action Via TGF-b on Osteoblasts 1005

PTH stimulated TGF-b1 mRNA expression inrat primary osteoblastic cells treated with ratPTH (1–34, 10�8M) for different times duringthe three stages of differentiation. The results(Fig. 4) indicate that PTH stimulated TGF-b1mRNA expression in differentiating andminer-alizing osteoblasts at 4 h treatment indicatingthat osteoblasts respond to PTH and synthesizeTGF-b1 only at these stages. Since PTH stimu-lates mRNA expression of both LTBP-1 andTGF-b1 mRNAs, we wanted to determine whe-ther there is a synergistic effect with combinedtreatment with PTH and TGF-b1 on LTBP-1expression in rat osteoblastic cells. UMR106-01cells were treated with human TGF-b1 (1 ng/ml), rat PTH (1–34, 10�8M) or both together atdifferent time periods. Total RNA was isolatedand subjected to real time RT-PCR analysis.

TGF-b1 and PTH stimulated LTBP-1 mRNAexpression to 3.5� 1.1-fold and 11.7� 2.1-fold,respectively at 4 h and to 1.4� 0.2-fold and7.3� 2.4-fold, respectively at 24 h in UMR 106-01 cells. When cells were treated with TGF-b1and PTH together, an additive effect wasobserved at both 4 h (17.1� 1.6-fold) and 24 h(9.5� 2.7-fold) in these cells (Fig. 5).

PTH Stimulation of LTBP-1 mRNA Expression IsDependent on the PKA Signaling Pathway

To identify the signaling pathways in PTH-stimulated LTBP-1 mRNA expression, we usedMAPK, PKA, and PKC pathway inhibitors.UMR 106-01 cells were pretreated with DMSO,PD98059 (MEK inhibitor), SB203580 (p38MAPK inhibitor), SP600125 (JNKII inhibitor),H89 (PKA inhibitor), or GF109203X (PKC

Fig. 2. Effect of PTH on expression of LTBP-1 mRNA levels inmouse preosteoblastic cells and in rat primary osteoblastic cells.A: MC3T3 cells were serum-starved for 24 h and treated withcontrol medium or medium containing rat PTH (1–34, 10�8M)for different time periods as indicated. Total RNA was isolatedand subjected to real time RT-PCR using specific primers for ratLTBP-1 and b-actin. B: Osteoblasts derived from postnatal day 1,rat calvariae were grown in 6-well plates in MEM, 10% FBS toconfluence (day 7), after which the cells were switched todifferentiation andmineralizingmedium (BGJb, 10% FBS, 50 mg/

ml ascorbic acid, and 10mM b-glycerophosphate). Proliferating,differentiating, and mineralizing osteoblasts were treated withcontrol medium or medium containing rat PTH (1–34, 10�8M)for 1, 4, and 12 h at days 7, 14, and 21 of culture. Total RNAwasisolated and subjected to real time RT-PCRusing specific primersfor rat LTBP-1 and b-actin. The relative levels of mRNAs werenormalized to b-actin, and the PTH-fold changes were calcu-lated over controls. The asterisks represent P<0.05 comparedwith control.

1006 Kwok et al.

inhibitor) for 20 min, then treated with orwithout rat PTH (1–34, 10�8M) for 4 h. TotalRNA was isolated and real time RT-PCR wasperformed. The PTH-stimulated LTBP-1mRNAexpression was not significantly decreased byMAPK and PKC inhibitors, suggesting that theMEK, p38 MAPK, JNK, and PKC signalingpathways are not involved in PTH stimulationof LTBP-1 mRNA expression in UMR 106-01cells. The PKA inhibitor, H89 inhibited PTH-stimulated LTBP-1 mRNA expression in thesecells (Fig. 6). The effective concentrations andspecificity of these inhibitors have been pre-viously determined [Selvamurugan et al., 2002;Selvamurugan et al., 2004].

PTH Stimulation of TGF-b1 mRNA ExpressionIs a Secondary Effect and Is Dependent

on the MAPK Signaling Pathway

Since PTH stimulatedTGF-b1mRNAexpres-sion in rat osteoblastic cells (Fig. 4), we wantedto determine whether this stimulation is aprimary effect and if it requires the PKAsignaling pathway as we found for PTH stimu-lation of LTBP-1 mRNA expression in UMR106-01 cells (Figs. 3A and 6). UMR 106-01 cellswere treated with control medium or mediumcontaining rat PTH (1–34, 10�8M) for 4 h in thepresence or absence of 30 mg/ml cycloheximideadded 1 h before treatment. Total RNA wassubjected to real time RT-PCR analysis usingspecific primers for rat TGF-b1 and b-actin. Asshown in Figure 7A, cycloheximide inhibitedPTH induction of TGF-b1 mRNA expression,indicating that PTH stimulation of TGF-b1expression is a secondary effect and de novoprotein synthesis is required for this purpose.To identify the signaling pathways in PTH-stimulated TGF-b1 mRNA expression, we usedMAPK, PKA, and PKC pathway inhibitors.Similar to Figure 6, UMR 106-01 cells werepretreated with DMSO, PD98059, SB203580,SP600125, H89, or GF109203X for 20min, thentreated with or without rat PTH (1–34, 10�8M)for 4 h. Total RNA was isolated and real timeRT-PCR was performed. The PTH stimulationof TGF-b1 mRNA expression was not signifi-cantly decreased by JNK, PKA, and PKCinhibitors; whereas MEK and p38 MAPKinhibitors inhibited PTH-stimulated TGF-b1mRNA expression in rat osteoblastic cells(Fig. 7).

DISCUSSION

The gene expression profile changes in UMR106-01 cells treated with rat PTH (1–34,10�8M) using DNAmicroarray analysis showedthat LTBP-1 is one of the genes stimulated bythis hormone [Qin et al., 2003]. We report herethat PTH stimulates LTBP-1mRNA expressionin rat osteoblastic and mouse preosteoblasticcells (Figs. 1 and 2). LTBP-1 is required for theproper folding and secretion of TGF-b [Miya-zono et al., 1991], which is a local factorproduced by both osteoblasts and osteoclasts[Pfeilschifter and Mundy, 1987; Bonewald andDallas, 1994]. Bone ECM is the major storagesite in the body for TGF-b [Seyedin et al., 1985;Hauschka et al., 1986]. This ECM-bound TGF-b,

Fig. 3. PTH-stimulated LTBP-1 mRNA expression is a primaryeffect. A: UMR 106-01 cells were serum-starved for 24 h andtreated with control medium or medium containing rat PTH (1–34, 10�8M) for 4 h in the presence or absence of 30 mg/mlcycloheximide (CHX) added 1 h before PTH treatment, and totalRNAwas subjected to real timeRT-PCRusing specific primers forrat LTBP-1 and b-actin.B: UMR106-01 cells were serum-starvedfor 24 h and treated with control medium or medium containingrat PTH (1–34, 10�8M) for 24 h in the presence or absence of30 mg/ml cycloheximide (CHX) added 1 h before PTH treatment,and total RNA was subjected to real time RT-PCR using specificprimers for ratMMP-13andb-actin. The relative levels ofmRNAswere normalized to b-actin, and the PTH-fold changes werecalculated over controls.

LTBP-1, a Mediator for PTH Action Via TGF-b on Osteoblasts 1007

which is predominantly the TGF-b1 isoform, isstored in a latent form and can be released andactivated by resorbing osteoclasts [Oreffo et al.,1989; Oursler, 1994]. Several mechanisms forthe activation of latent TGF-b complexes havebeenwell documented [Munger et al., 1997; Koliet al., 2001].

TGF-b can influence many of the steps in thebone remodeling pathway. It can both inhibit[Hughes et al., 1996] and stimulate osteoclast

activity [Horwoodetal., 1999;SellsGalvin etal.,1999] depending on conditions. TGF-b inhibitsosteoclast activity, both by stimulating osteo-clasts to undergo apoptosis and by inhibitingformation of osteoclasts from their precursors.TGF-b is also a powerful chemoattractant andmitogen for osteoblast precursors [Bonewald,1999]. The effect ofTGF-b onmature osteoblastsis then to inhibit proliferation and stimulateproduction of bone ECM proteins, including

Fig. 4. Effect of PTH on expression of TGF-b1 mRNA levels inrat primary osteoblastic cells. Proliferating, differentiating, andmineralizing rat primary osteoblasts were treated with controlmedium or medium containing rat PTH (1–34, 10�8M) for 1, 4,and 12 h at days 7, 14, and 21 of culture. Total RNAwas isolated

and subjected to real time RT-PCR using specific primers forrat TGF-b1 and b-actin. The relative levels of mRNAs werenormalized to b-actin, and the PTH-fold changes were calcu-lated over controls. The asterisks represent P<0.05 comparedwith control.

Fig. 5. Effect of PTH and TGF-b1 on expression of LTBP-1mRNA levels in rat osteoblastic cells. UMR 106-01 cells wereserum-starved for 24 h and treated with control medium ormedium containing rat PTH (1–34, 10�8M) (P), human TGF-b1(1 ng/ml) (T), or both together for 1, 4, and 24 h. Total RNA wasisolated and subjected to real timeRT-PCRusing specific primersfor rat LTBP-1 and b-actin. The relative levels of mRNAs werenormalized to b-actin, and the PTH-fold changes were calcu-lated over controls. The asterisks represent P< 0.05 comparedwith control.

Fig. 6. PTH-stimulated LTBP-1 mRNA expression depends onthe PKA signaling pathway. UMR 106-01 cells were serum-starved for 24 h and then treated with control or rat PTH (1–34,10�8M)-containing medium for 4 h in the presence or absenceof inhibitors PD98059, SB203580, SP600125, H89, andGF109203X (added 20min before PTH). Total RNAwas isolatedand subjected to real time RT-PCR using specific primers for ratLTBP-1 and b-actin. The relative levels of mRNAs were normal-ized to b-actin, and the PTH-fold changes were calculated overcontrols. The asterisks representP< 0.05 comparedwith control.

1008 Kwok et al.

type I collagen, fibronectin, and osteocalcin[Roberts, 1998]. TGF-b has therefore beenimplicated as a coupling factor that coordinatesthe processes of bone resorption and subsequentbone formation. A functional role for LTBP inregulating the local activity of TGF-b hasemerged using antibodies to LTBP-1 [Miyazonoet al., 1991]. TGF-b1 also induces expression ofits own mRNA as well as expression of LTBP-1[Dallas et al., 1994; Roberts, 1998]. This isconsistent with our results that TGF-b1 stimu-lates LTBP-1 mRNA expression in rat primaryosteoblastic cells (Fig. 4). We report here thatPTH stimulated expression of both LTBP-1 andTGF-b1 in rat osteoblastic cells and this effectmay be required to maintain the level of LTBP

and its bound latent TGF-b in bone matrix. It ismost likely that the PTH stimulation of LTBP-1expression in osteoblasts has an effect on osteo-clasts via TGF-b. It is possible that LTBP-1mayplay an important role to link signaling betweenthe systemic (PTH) and local (TGF-b) factors.This may be one of the PTH regulatory mechan-isms that is necessary for maintaining thebalance between osteoblastic and osteoclasticactivity.

The PTH effect on LTBP-1 expression is aprimary effect thus, not requiring de novoprotein synthesis (Fig. 3A). This result supportsthe fact that, by association of TGF-b with theECM, it is stored in a readily mobilized form,which could allow extracellular signaling toproceed rapidly in the absence of new proteinsynthesis. This event is particularly importantin situations such as tissue repair followinginjury. PTH and TGF-b1 stimulated LTBP-1mRNAexpression (Fig. 5) andboth together hadan additive effect on LTBP-1mRNA expression,indicating that PTH and TGF-b may haveseparate intracellular components to activateLTBP-1 gene expression. It is well documentedthat PTH mediates its effects by the PTH1R,and TGF-bmediates its effects by TGF-b type IIand type I receptors [Attisano andWrana, 1998;Massague and Wotton, 2000; Swarthout et al.,2002; Derynck and Zhang, 2003; Qin et al.,2004]. Even though PTH activates both PKAand PKC signaling pathways [Swarthout et al.,2002; Qin et al., 2004], we identified that thePKA signaling pathway is responsible for PTH-stimulated LTBP-1 mRNA expression in UMR106-01 cells (Fig. 6). The requirement of thePKA signaling pathway for PTH stimulation ofLTBP-1 mRNA expression (Fig. 6) and therequirement of de novo protein synthesis andtheMEKandp38MAPKsignalingpathways forPTH-stimulated TGF-b1 mRNA expression(Fig. 7) suggest that PTH mediates its effectson osteoblasts by several intracellular signalingcomponents in rat osteoblastic cells. In sum-mary, we provide evidence for PTH stimulatedand PKA-dependent LTBP-1 mRNA expressionin osteoblastic cells, which appears to beimportant for PTH regulation of the TGF-bsystem to mediate bone remodeling activities.

ACKNOWLEDGMENTS

This research was supported by grants fromthe Department of Defense (W81XWH-04-1-

Fig. 7. PTH-stimulated TGF-b1 mRNA expression requires denovo protein synthesis and depends on the MEK and p38 MAPKsignaling pathways.A: UMR106-01 cellswere serum-starved for24 h and treated with control medium or medium containing ratPTH (1–34, 10�8M) for 4 h in the presence or absence of 30 mg/ml cycloheximide (CHX) added 1 h before PTH treatment, andtotal RNA was subjected to real time RT-PCR using specificprimers for rat TGF-b1 and b-actin. B: UMR 106-01 cells wereserum-starved for 24 h and then treated with control or rat PTH(1–34, 10�8M)-containing medium for 4 h in the presence orabsence of inhibitors PD98059, SB203580, SP600125,H89, andGF109203X (added 20min before PTH). Total RNAwas isolatedand subjected to real time RT-PCR using specific primers forrat TGF-b1 and b-actin. The relative levels of mRNAs werenormalized to b-actin, and the PTH-fold changes were calcu-lated over controls. The asterisks represent P< 0.05 comparedwith control.

LTBP-1, a Mediator for PTH Action Via TGF-b on Osteoblasts 1009

0687), the New Jersey Commission on CancerResearch and the Foundation of the Universityof Medicine and Dentistry of New Jersey (toN.S.) and by NIH grants DK48109 andDK47420 (to N.C.P.).

REFERENCES

Annes JP, Munger JS, Rifkin DB. 2003. Making sense oflatent TGFbeta activation. J Cell Sci 116:217–224.

Attisano L, Wrana JL. 1998. Mads and Smads in TGF betasignalling. Curr Opin Cell Biol 10:188–194.

Bellows CG, Ishida H, Aubin JE, Heersche JN. 1990.Parathyroid hormone reversibly suppresses the differ-entiation of osteoprogenitor cells into functional osteo-blasts. Endocrinology 127:3111–3116.

Bonewald LF. 1999. Regulation and regulatory activities oftransforming growth factor beta. Crit Rev EukaryotGene Expr 9:33–44.

Bonewald LF, Dallas SL. 1994. Role of active and latenttransforming growth factor beta in bone formation. J CellBiochem 55:350–357.

Bonewald LF, Wakefield L, Oreffo RO, Escobedo A,Twardzik DR, Mundy GR. 1991. Latent forms oftransforming growth factor-beta (TGF beta) derived frombone cultures: Identification of a naturally occurring100-kDa complex with similarity to recombinant latentTGF beta. Mol Endocrinol 5:741–751.

Dallas SL, Park-Snyder S, Miyazono K, Twardzik D,Mundy GR, Bonewald LF. 1994. Characterization andautoregulation of latent transforming growth factorbeta (TGF beta) complexes in osteoblast-like celllines. Production of a latent complex lacking the latentTGF beta-binding protein. J Biol Chem 269:6815–6821.

Dallas SL, Miyazono K, Skerry TM, Mundy GR, BonewaldLF. 1995. Dual role for the latent transforming growthfactor-beta binding protein in storage of latent TGF-betain the extracellular matrix and as a structural matrixprotein. J Cell Biol 131:539–549.

Dallas SL, Keene DR, Bruder SP, Saharinen J, Sakai LY,Mundy GR, Bonewald LF. 2000. Role of the latenttransforming growth factor beta binding protein 1 infibrillin-containing microfibrils in bone cells in vitro andin vivo. J Bone Miner Res 15:68–81.

Dallas SL, Zhao S, Cramer SD, Chen Z, Peehl DM,Bonewald LF. 2005. Preferential production of latenttransforming growth factor beta-2 by primary prostaticepithelial cells and its activation by prostate-specificantigen. J Cell Physiol 202:361–370.

Derynck R, Zhang YE. 2003. Smad-dependent and Smad-independent pathways in TGF-beta family signalling.Nature 425:577–584.

Dobnig H, Turner RT. 1997. The effects of programmedadministration of human parathyroid hormone fragment(1–34) on bone histomorphometry and serum chemistryin rats. Endocrinology 138:4607–4612.

Hauschka PV, Mavrakos AE, Iafrati MD, Doleman SE,Klagsbrun M. 1986. Growth factors in bone matrix.Isolation of multiple types by affinity chromatography onheparin-Sepharose. J Biol Chem 261:12665–12674.

HorwoodNJ, Kartsogiannis V, Quinn JM, Romas E,MartinTJ, Gillespie MT. 1999. Activated T lymphocytes support

osteoclast formation in vitro. Biochem Biophys ResCommunun 265:144–150.

Hughes DE, Dai A, Tiffee JC, Li HH, Mundy GR, BoyceBF. 1996. Estrogen promotes apoptosis of murine osteo-clasts mediated by TGF-beta. Nature Med 2:1132–1136.

Kanzaki T, Olofsson A, Moren A, Wernstedt C, Hellman U,Miyazono K, Claesson-Welsh L, Heldin CH. 1990. TGF-beta 1 binding protein: A component of the large latentcomplex of TGF-beta 1 with multiple repeat sequences.Cell 61:1051–1061.

Kanzawa M, Sugimoto T, Kanatani M, Chihara K. 2000.Involvement of osteoprotegerin/osteoclastogenesis inhi-bitory factor in the stimulation of osteoclast formation byparathyroid hormone in mouse bone cells. Eur J Endo-crinol 142:661–664.

Koli K, Saharinen J, Karkkainen M, Keski-Oja J. 2001.Novel non-TGF-beta-binding splice variant of LTBP-4 inhuman cells and tissues providesmeans to decrease TGF-beta deposition. J Cell Sci 114:2869–2878.

Koski C, Saharinen J, Keski-Oja J. 1999. J Biol Chem274:32619–32630.

Marra F, Bonewald LF, Park-Snyder S, Park IS, WoodruffKA, Abboud HE. 1996. Characterization and regulationof the latent transforming growth factor-beta complexsecreted by vascular pericytes. J Cell Physiol 166:537–546.

Massague J, Wotton D. 2000. Transcriptional control by theTGF-beta/Smadsignaling system.EMBOJ19:1745–1754.

Miyazono K, Olofsson A, Colosetti P, Heldin CH. 1991. Arole of the latent TGF-beta 1-binding protein in the as-sembly and secretion of TGF-beta 1. EMBO J 10:1091–1101.

Munger JS, Harpel JG, Gleizes PE, Mazzieri R, Nunes I,Rifkin DB. 1997. Latent transforming growth factor-beta: Structural features and mechanisms of activation.Kidney Int 51:1376–1382.

Oklu R, Hesketh R. 2000. The latent transforming growthfactor beta binding protein (LTBP) family. Biochem J352:601–610.

Oreffo RO, Mundy GR, Seyedin SM, Bonewald LF. 1989.Activation of the bone-derived latent TGF beta complexby isolated osteoclasts. Biochem Biophys Research Com-mun 158:817–823.

Oursler MJ. 1994. Osteoclast synthesis and secretion andactivation of latent transforming growth factor beta.J Bone Miner Res 9:443–452.

Pedrozo HA, Schwartz Z, Robinson M, Gomes R, Dean DD,Bonewald LF, Boyan BD. 1999. Potential mechanismsfor the plasmin-mediated release and activation of latenttransforming growth factor-beta1 from the extracellularmatrix of growth plate chondrocytes. Endocrinology 140:5806–5816.

Pfeilschifter J, Mundy GR. 1987. Modulation of type betatransforming growth factor activity in bone cultures byosteotropic hormones. Proc Natl Acad Sci USA 84:2024–2028.

Qin L, Qiu P, Wang L, Li X, Swarthout JT, Steropoulos P,Tolias P, Partridge NC. 2003. Gene expression profilesand transcription factors involved in parathyroid hor-mone signaling in osteoblasts revealed bymicroarray andbioinformatics. J Biol Chem 278:19723–19731.

Qin L, Raggatt LJ, Partridge NC. 2004. Parathyroidhormone: A double-edged sword for bone metabolism.Trends Endocrinol Metab 15:60–65.

1010 Kwok et al.

Ramirez F, Pereira L. 1999. The fibrillins. Int J BiochemCell Biol 31:255–259.

Rifkin DB. 2005. Latent TGF-beta binding proteins:Orchestrators of TGF-beta availability. J Biol Chem280:7409–7412.

Roberts AB. 1998. Molecular and cell biology of TGF-beta.Miner Electrolyte Metab 24:111–119.

Saharinen J, Hyytiainen M, Taipale J, Keski-Oja J. 1999.Latent transforming growth factor-beta binding proteins(LTBPs)—structural extracellular matrix proteins fortargeting TGF-beta action. Cytokine Growth Factor Rev10:99–117.

Scott DK, Brakenhoff KD, Clohisy JC, Quinn CO, PartridgeNC. 1992. Parathyroid hormone induces transcription ofcollagenase in rat osteoblastic cells by amechanism usingcyclic adenosine 30,50-monophosphate and requiringprotein synthesis. Mol Endocrinol 6:2153–2159.

Sells Galvin RJ, Gatlin CL, Horn JW, Fuson TR. 1999.TGF-beta enhances osteoclast differentiation in hemato-poietic cell cultures stimulated with RANKL andM-CSF.Biochem Biophys Res Commun 265:233–239.

Selvamurugan N, Fung Z, Partridge NC. 2002. Transcrip-tional activation of collagenase-3 by transforming growthfactor-beta1 is via MAPK and Smad pathways in humanbreast cancer cells. FEBS Lett 532:31–35.

Selvamurugan N, Kwok S, Alliston T, Reiss M, PartridgeNC. 2004. Transforming growth factor-beta 1 regulationof collagenase-3 expression in osteoblastic cells by cross-talk between the Smad and MAPK signaling pathwaysand their components, Smad2 and Runx2. J Biol Chem279:19327–19334.

Seyedin SM, Thomas TC, Thompson AY, Rosen DM, PiezKA. 1985. Purification and characterization of twocartilage-inducing factors from bovine demineralizedbone. Proc Natl Acad Sci USA 82:2267–2271.

Shalhoub V, Conlon D, Tassinari M, Quinn C, Partridge N,Stein GS, Lian JB. 1992. Glucocorticoids promote devel-opment of the osteoblast phenotype by selectivelymodulating expression of cell growth and differentiationassociated genes. J Cell Biochem 50:425–440.

Swarthout JT, D’Alonzo RC, Selvamurugan N, PartridgeNC. 2002. Parathyroid hormone-dependent signalingpathways regulating genes in bone cells. Gene 282:1–17.

Taipale J, Miyazono K, Heldin CH, Keski-Oja J. 1994.Latent transforming growth factor-beta 1 associates tofibroblast extracellular matrix via latent TGF-betabinding protein. J Cell Biol 124:171–181.

Tam CS, Heersche JN, Murray TM, Parsons JA. 1982.Parathyroid hormone stimulates the bone apposition rateindependently of its resorptive action: Differential effectsof intermittent and continuous administration. Endocri-nology 110:506–512.

Wakefield LM, Smith DM, Flanders KC, Sporn MB. 1988.Latent transforming growth factor-beta from humanplatelets. A high molecular weight complex containingprecursor sequences. J Biol Chem 263:7646–7654.

Winchester SK, Bloch SR, Fiacco GJ, Partridge NC.1999. Regulation of expression of collagenase-3 innormal, differentiating rat osteoblasts. J Cell Physiol181:479–488.

LTBP-1, a Mediator for PTH Action Via TGF-b on Osteoblasts 1011

Author Proof

AJournal of Cellular Biochemistry 9999:1–12 (2006)

Overexpression of Runx2 Directed by the MatrixMetalloproteinase-13 Promoter Containing the AP-1 andRunx/RD/Cbfa Sites Alters Bone Remodeling In Vivo

Nagarajan Selvamurugan,1 Stephen C. Jefcoat,1 Sukyee Kwok,1 Yibing Yang,1

Rodney Kowalewski,1 Joseph A. Tamasi,2 and Nicola C. Partridge1*1Department of Physiology and Biophysics, UMDNJ-Robert Wood Johnson Medical School,Piscataway, New Jersey 088542Osteoporosis Research, Metabolic and Cardiovascular Drug Discovery PRI,Bristol-Myers Squibb Company, Pennington, New Jersey 08534

Abstract The activator protein-1 (AP-1) and runt domain binding (Runx/RD/Cbfa) sites and their respective bindingproteins, c-Fos/c-Jun and Runx2 (Cbfa1), regulate the rat matrix metalloproteinase-13 (MMP-13) promoter in bothparathyroid hormone (PTH)-treated and differentiating osteoblastic cells in culture. To determine the importance of theseregulatory sites in the expression of MMP-13 in vivo, transgenic mice containing either wild-type (�456 or�148) or AP-1and Runx/RD/Cbfa sites mutated (�148A3R3) MMP-13 promoters fused with the E. coli lacZ reporter were generated. Thewild-type transgenic lines expressed higher levels of bacterial b-galactosidase in bone, teeth, and skin compared to themutant and non-transgenic lines. Next, we investigated if overexpression of Runx2 directed by the MMP-13 promoterregulated expression of bone specific genes in vivo, and whether this causes morphological changes in these animals. Realtime RT-PCR experiments identified increased mRNA expression of bone forming genes and decreased MMP-13 in thetibiae of transgenic mice (14 days and 6 weeks old). Histomorphometric analyses of the proximal tibiae showed increasedbone mineralization surface, mineral apposition rate, and bone formation rate in the transgenic mice which appears to bedue to decreased osteoclast number. Since MMP-13 is likely to play a role in recruiting osteoclasts to the bone surface,decreased expression of MMP-13 may cause reduced osteoclast-mediated bone resorption, resulting in greater boneformation in transgenic mice. In summary, we show here that the 148 bp upstream of the MMP-13 transcriptional start siteis sufficient and necessary for gene expression in bone, teeth, and skin in vivo and the AP-1 and Runx/RD/Cbfa sitesare likely to regulate this. Overexpression of Runx2 by these regulatory elements appears to alter the balance between thebone formation-bone resorption processes in vivo. J. Cell. Biochem. 9999: 1–12, 2006. � 2006 Wiley-Liss, Inc.

Key words: bone formation; bone remodeling; bone resorption; Runx2; collagenase-3; matrix metalloproteinase-13

Bone is the body’s main reservoir of calciumand phosphate ions, and is constantly regener-ated through continuous formation and resorp-tion in the process of bone remodeling. Thisphysiological process occurs throughout adult

life to maintain a constant bone mass. In severalpathological conditions, the tight balancebetween bone formation and resorption is notpreserved [Avioli and Krane, 1998]. Boneformation in vivo is a complex phenomenon

JCB-05-0464(20878)

� 2006 Wiley-Liss, Inc.

*Correspondence to: Nicola C. Partridge, PhD, Departmentof Physiology and Biophysics, UMDNJ-Robert Wood John-son Medical School, 675 Hoes Lane, Piscataway, NJ 08854.E-mail: partrinc@umdnj.edu

Received 20 October 2005; Accepted 1 February 2006

DOI 10.1002/jcb.20878Published online 00 Month 2006 in Wiley InterScience(www.interscience.wiley.com).

Abbreviations used: MMPs, matrix metalloproteinases;AP-1, activator protein-1; RD, runt domain binding site;Cbfa, core binding factor alpha; AML, acute myeloid leuke-mia; ECM, extracellular matrix; PTH, parathyroid hormone;RT-PCR, reverse transcriptase-polymerase chain reaction;BSP, bone sialoprotein; OC, osteocalcin; ALP, alkalinephosphatase; OPN, osteopontin; OPG, osteoprotegerin;RANKL, receptor activator of nuclear factor kappaB ligand.

Grant sponsor: National Institutes of Health; Grantnumber: DK47420 and DK48109; Grant sponsor: NewJersey Commission on Cancer Research and the Depart-ment of Defense, US Army; Grant number: W81XWH-04-1-0687.

Author Proof

Awhereby recruitment and replication of mesen-chymal precursors of osteoblasts, differentiat-ion into preosteoblasts, osteoblasts, and matureosteoblasts ultimately results in the accumula-tion and mineralization of the extracellularmatrix [Aubin, 1998]. Osteoclasts responsiblefor bone resorption are derived from hematopoie-tic precursor cells belonging to the monocyte/macrophage lineage. Osteoclast differentiationis a multistep process that leads eventuallyto multinucleated bone-resorbing osteoclasts[Roodman, 1999].

Runx2, also called Cbfa1 or Pebp2aA, is atranscription factor that belongs to the runt-domain gene family [Komori and Kishimoto,1998]. Runt proteins are a group of trans-cription factors conserved from C. elegans tohumans. They share a typical 128-amino acidDNA-binding domain called the Runt domain.Runx2 acts as an inducer of osteoblast differ-entiation and it has all the characteristics of adifferentiation regulator in the osteoblast line-age [Ducy et al., 1997; Komori et al., 1997;Mundlos et al., 1997; Otto et al., 1997; Karsenty,1999]. Bone formation is not solely controlled atthe level of osteoblast differentiation and Runx2is also required for osteoblast function [Ducyet al., 1999]. Runx2 is able to induce both earlyand late markers for osteoblast differentiation,including alkaline phosphatase (ALP), type Icollagen, osteopontin (OPN), bone sialoprotein,and osteocalcin (OC) in several cell lines[Harada et al., 1999]. Genes involved in thebone resorption process, such as RANKL, OPG,and MMP-13 are also regulated by Runx2[Geoffroy et al., 2002].

MMP-13, a matrix metalloproteinase, isexpressed as a late-differentiation gene in osteo-blasts, and is primarily responsible for thedegradation of extracellular bone matrix com-ponents (type I, II, and III fibrillar collagens).MMP-13 gene expression is regulated by bone-resorbing agents, such as PTH, cytokines suchas interleukin-1 and -6, and growth factorsthat promote bone turnover [Scott et al., 1992;Varghese et al., 1995, 1996, 2000; Kusano et al.,1998]. In vivo, MMP-13 has been shown to beexpressed in ossifying centers during bonedevelopment [Schorpp et al., 1995] and is detec-table by immunohistochemistry in rat calvariae14 days after birth [Partridge et al., 1998]. Theregulation of this gene is likely to have impor-tant consequences for both normal and patho-logical remodeling of bone where the balance

between bone resorption and bone formation isdisrupted. Using mutant mice homozygous for atargeted mutation in Col1a1 that are resistantto collagenase cleavage of type I collagen, Zhaoet al. [1999] showed that PTH-induced boneresorption and calcemic responses were mark-edly diminished. The number of osteoclasts wasalso reduced and the animals had thicker thannormal bones [Zhao et al., 2000]. Studies with anull mutation of the MMP-13 gene in miceshowed that in Mmp13�/� embryos, the growthplates were strikingly lengthened, a defectascribable predominantly to a delay in terminalevents in the growth plates, with failure toresorb collagens, as well as a delay in ossifica-tion at the primary centers [Inada et al., 2004].

Previous work in our laboratory has deter-mined that the activator protein-1 (AP-1) andrunt domain binding (Runx/RD/Cbfa) sites andtheir respective binding proteins, c-Fos/c-Junand Runx2 (Cbfa1), regulate the MMP-13promoter in both PTH-treated and differentiat-ing osteoblastic cells in culture [Selvamuruganet al., 1998; Winchester et al., 2000]. Further-more, protein–protein interaction studies indi-cate that Runx2 and the runt domain of Runx2alone interact with c-Fos and c-Jun [D’Alonzoet al., 2002]. Also, co-transfection of Runx2 withthe MMP-13 promoter in UMR 106-01 cells hasbeen shown to enhance transactivation of theMMP-13 promoter [D’Alonzo et al., 2002]. Inthe present study, we wished to investigate theimportance of the AP-1 and Runx/RD/Cbfa sitesand Runx2 in the expression of MMP-13 andbone specific genes in vivo. To determine if theregulatory elements responsible for MMP-13expression in vitro operated in vivo, transgenicmice containing either wild-type (�456 or�148)or AP-1 and Runx/RD/Cbfa site mutant(�148A3R3) MMP-13 promoters fused withthe E. coli lacZ (b-galactosidase) reporter weregenerated. Since Runx2 is involved in bonedevelopment and bone remodeling, we alsoinvestigated if overexpression of Runx2 underthe control of the MMP-13 promoter in vivoregulates bone specific genes, and whether thiscauses changes in the bone phenotype of theseanimals.

MATERIALS AND METHODS

Transgenic Mice Generation

The rat MMP-13 promoter fragments weregenerated by PCR, with SalI linkers engineered

2 Selvamurugan et al.

Author Proof

Aonto the 50 and 30 ends. The fragments weresubcloned into pSV0CAT (Promega, Madison,WI) vectors and the AP-1 and RD mutationswere generated using the Chameleon double-stranded site-directed mutagenesis kit (Strata-gene). The fragments were released by SalIdigestion, ligated to bacterial b-galactosidaseand mouse protamine 1 which is a proteinnecessary for haploid DNA packaging andpaternal procreation. These transgenes werethen gel purified and used for injection intomouse blastocysts. The pCMV-c-myc mamma-lian expression vector expressing the N-term-inal c-myc epitope tag was used to clone the ratMMP-13 promoter (�148 containing the AP-1and RD sites) driving Runx2 (type II). Theentire fragment containing the rat MMP-13promoter, c-myc epitope tag, and Runx2 wasreleased by a SphI and PvuII digestion and usedfor injection into mouse blastocysts. The Insti-tutional Animal Care and Use Committee atSaint Louis University approved all proceduresfor the generation of the mice and collection oftissues from the mice bearing the MMP-13promoter/reporter genes and the MMP-13 pro-moter/Runx2 genes. The Institutional AnimalCare and Use Committee of UMDNJ-RobertWood Johnson Medical School approved alltreatments and procedures for collection oftissues from both sets of mice.

RNA Isolation and Real Time RT-PCR

Bone samples were ground with the grindingmill in a TRIZOL (Invitrogen) solution. RNAextraction was performed according to theinstructions given by the company. Reversetranscriptase (RT) reaction was carried outusing the TaqMan Reverse Transcription rea-gents (Roche). PCR reactions were performedaccording to the real-time thermocycler machinemanufacturer’s instructions (DNA EngineOpticon, MJ Research, MA), which allow real-time quantitative detection of the PCR productby measuring the increase in SYBR greenfluorescence caused by binding of SYBR greento double-stranded DNA. The SYBR green kitfor PCR reactions was purchased from PerkinElmer Applied Biosystems. Primers used in thisstudy were designed using the PrimerExpresssoftware (Perkin Elmer Applied Biosystems).For PCR amplification, the following sets ofprimers were used: OC, 50 AAGCAGGAGGG-CAATAAGGT 30 and 50 AGCTGCTGTGACAT-CCCATAC 30; MMP-13, 50 GCCACCTTCTT-

CTTGTTGAGCTG 30 and 50 ATCAAGGGATA-GGGCTGGGTCAC 30; ALP, 50 AGGCAGGATT-GACCACGG 30 and 50 TGTAGTTCTGCTCAT-GGA 30; OPN, 50 CCAATGAAAGCCATGAC-CACA 30 and 50 CGTCAGATTCATCCGAGTC-CAC 30; Osteoprotegerin (OPG), 50 CGAGGA-CCACAATGAACAAG 30 and 50 TCTCGGCATT-CACTTTGGTC 30; RANKL, 50 CAGAAGACAG-CACTCACTGC 30 and 50 ATGGGAACCCGAT-GGGATGC 30.

Immunohistochemistry

Soft tissue samples were harvested and fixedin fresh 4% paraformaldehyde in PBS at 48C for24 h. Bone tissue was harvested and immedi-ately placed into fresh 4% paraformaldehydeand incubated at 48C for 1 h. The bones weredemineralized in 0.1 M Tris-HCl, 0.3 M EDTA,pH 7.4 at 48C for 4 days with daily changes.Fixed samples were embedded in cryopreserva-tive (O.C.T. Compound), frozen in nitrogen-cooled isopentane and stored at �708C untilsectioning. Five-micrometer frozen sectionswere thawed at room temperature, and washedin PBS twice, 5 min each. The sections werepostfixed in 4% paraformaldehyde for 15 min,and after rinsing with PBS/BSA (1 mg/ml), werekept for 30 min in 0.3% H2O2 in methanol toblock endogenous peroxidase activity. Afterbrief rinsing, non-specific antibody bindingwas suppressed by normal rabbit serum diluted1:70 in PBS/BSA for 30 min. After a brief rinse,the sections were incubated with polyclonalrabbit anti-bacterial b-galactosidase antibody(Rockland, Inc., Gilbertsville, PA) at 2 mg/ml(1:500 dilution) in PBS/BSA at 48C overnight.After rinsing, the sections were incubatedwith a biotinylated anti-rabbit antibody. Thebound antibody complex was visualized bythe avidin-biotin-peroxidase procedure usingthe Vectastain ABC Elite Kit (Vector Labora-tories, Burlingame, CA) and 3-amino-9-ethyl-carbazole (AEC) as chromogen. Control sectionsincubated with biotinylated rabbit IgG or sec-tions from non-transgenic animals showed verylow background and no specific cell staining.Photomicrographs were taken with the use of aSpot insight color digital camera (DiagnosticInstruments, Inc., Sterling Heights, MI) attach-ed to a Nikon Microphot-FXL microscope (Nikon,Melville, NY), and Spot Advanced imagingsoftware (Diagnostic Instruments, Inc., Ster-ling Heights).

MMP-13 Promoter and Runx2 in Mice 3

Author Proof

AHistology

Bones from the wild-type mice and transgenicmice were decalcified and fixed in 4% parafor-maldehyde/0.1 M phosphate buffer. They werethen paraffin-embedded, sectioned (5 mm thick),and stained with hematoxylin and eosin.

mCT (Microcomputed Tomography)

Measurements of trabecular architecturewere done on the proximal tibiae cleared of allsoft tissue using a mCT 20 (Scanco Medical AG,Bassersdorf, Switzerland). After an initial scoutscan, a total of 100 slices with an increment of22mm were obtained on each bone sample, start-ing 1.5 mm below the growth plate in the area ofthe secondary spongiosa. The area for analysiswas outlined within the trabecular compart-ment, excluding the cortical and subcorticalbone. Every 10 sections were outlined, and theintermediate sections were interpolated withthe contouring algorithm to create a volume ofinterest. Segmentation values used for analysiswere sigma 1.2, support 2, and threshold 286. Athree-dimensional (3-D) analysis was done todetermine bone volume (BV/TV), trabecularnumber (Tb.N), trabecular thickness (Tb.Th),trabecular separation (Tb.Sp), and connectivitydensity (Conn.D). Cross-sectional area wasdetermined by outlining the periosteal surfaceand performing a two-dimensional analysis.

Cortical bone was measured on the tibia 2 mmbelow the tibia-fibula junction, where the dia-physis is most uniform in shape. Ten slices of thediaphysis were made, and the same segmenta-tion parameters were used for analysis. Theperiosteal surface was outlined, and a two-dimensional analysis was done to determinecross-sectional area, bone volume, and perios-teal perimeter (Ps.Pm). The endocortical sur-face was outlined, and the analysis repeated todetermine endocortical perimeter (Ec.Pm). Themean cortical thickness (Ct.Th) was determinedby distance measurements at eight differentpoints on the cortical slice.

Histomorphometric Analyses

For the assessment of dynamic histomorpho-metric indices, mice were injected twice withcalcein at a dose of 16 mg per g body weightand analyzed at 14 days or 6 weeks of age.The 14 days group received dual injections at 7and 2 days before sacrifice, and the 6-weekgroup received them at 12 and 2 days before

sacrifice. Tibiae were fixed with ethanol, andthe calcified bones were embedded in glycol-methacrylate. Three-micrometer longitudinalsections from the proximal parts of tibiae werestained with toluidine blue and analyzed usinga semiautomated system (Osteoplan II; ZEISS).Nomenclature, symbols, and units used arethose recommended by the Nomenclature Com-mittee of the American Society for Bone andMineral Research [Parfitt et al., 1987]. Thehistomorphometric service was provided by theCenter for Metabolic Bone Disease, Universityof Alabama, Birmingham, AL.

Statistical Analysis

The results are expressed as means�standard errors of the means (SEM). Statisticalanalyses were carried out using Student’s t test(MicroSoft Excel 97). All statistical tests weretwo tailed and unpaired.

RESULTS

We have previously shown that 148 bp up-stream of the transcriptional start site of the ratMMP-13 promoter retains all PTH-responsive-ness and differentiating elements in osteoblas-tic cells [Selvamurugan et al., 1998; Winchesteret al., 2000]. Three transgenic mouse lines weregenerated which carried the E. coli b-galactosi-dase reporter (marker) gene attached to 148 bpsequence upstream of the MMP-13 gene (Fig. 1).Transgenic founder mice were identified byPCR and Southern blot analyses of DNA iso-lated from tail snips. The bone and soft tissues

Fig. 1. Generation of transgenic mice overexpressing E. colilacZ (b-galactosidase). The structure of the rat MMP-13 promoterfragments and nlacZ (nuclear lacZ) hybrid genes that weremicroinjected into mouse blastocysts.

4 Selvamurugan et al.

Author Proof

Afrom transgenic mice and non-transgenicsiblings at various ages were analyzed forexpression of b-galactosidase by immunohisto-chemical staining using anti-b-galactosidaseantibody. Expression could be detected in all ofthe lines at 14–17 days postnatally in bone(Fig. 2) and teeth (data not shown). Some signalwas also seen in skin (data not shown). No signalwas seen in other tissues. Faint staining fromexpression of the transgene (�148 promoter/b-galactosidase) can be observed in developingcartilage in the 14.5-day embryo, which is moreevident in the cartilage of the spine in 17.5-dayembryos. There is also some expression of thistransgene in the skin of postnatal animals (datanot shown). Thus, it appears that expression ofMMP-13 gene in bone, teeth, and skin is regu-lated by elements in the 148 bp upstream of thetranscription start site.

To be sure that no other elements are involvedin tissue-specific expression in vivo, transgenicmice harboring �456 bp upstream sequence ofthe rat MMP-13 gene attached to the E. coli b-galactosidase marker gene were also generated(Fig. 1). Two transgenic founders were identi-fied by PCR of genomic DNA obtained from tailbiopsies. Similar experiments were conductedas for the �148/b-galactosidase construct. Wehave found that the transgene is expressed inbone at 14–17 days postnatally. The level ofexpression of the reporter gene seems greaterwith the �456 bp promoter compared with the�148 bp promoter. This may be due to otherelements such as Smad Binding Elements(SBEs) in the �456 to �148 region. Tissuessuch as heart, liver, and lung show no expres-sion of either of these constructs (Fig. 2). Wehave also generated three lines of mice with theRunx/RD/Cbfa and AP-1 sites mutated (Fig. 1)similar to our in vitro experiments [Selvamur-ugan et al., 1998; Winchester et al., 2000; Inadaet al., 2004]. The �148 construct with the RDand AP-1 site mutated (�148A3R3) showedexpression but it is far less than the wild-type�148 or �456 constructs at 14 days postnatalin calvariae (Fig. 2). It appears that 148 bpupstream of the transcriptional start site of therat MMP-13 promoter is sufficient to confergene expression in bone, teeth, and skin.

Runx2 is a bone specific transcription factorand is able to induce expression of genes that areinvolved in bone formation and bone resorption[Ducy et al., 1997; Harada et al., 1999; Liu et al.,2001; Geoffroy et al., 2002]. We wanted to

investigate whether overexpression of Runx2by the MMP-13 promoter regulates expressionof bone specific and bone related genes (MMP-13) in vivo, and whether this causes morpholo-gical changes in these animals. The pCMV-mycmammalian expression vector (Clontech) exp-ressing the N-terminal c-myc epitope tag wasused to clone both the rat MMP-13 promoter(�148 containing the AP-1 and RD sites) andRunx2 (type II). The entire fragment containingthe rat MMP-13 promoter, c-myc epitope tag,and Runx2 was released by SphI and PvuIIdigestion (Fig. 3A) and used for injection intomice blastocysts for the transgenic work. Todetermine that the transgene is expressed andis regulated by PTH, the transgenic DNAconstruct (�148/c-myc/Runx2), the negativecontrol constructs (pSV0/-148, pCMV/c-myc),and the positive control (pCMV/c-myc/Runx2)were transiently transfected into COS-7 cellsfor 48 h and then treated with or without8-bromo-cAMP (8BrcAMP) for 24 h. The cellswere then lysed and the c-myc-tagged Runx2(Cbfa1) was identified by Western blot using thec-myc antibody (Fig. 3B). The result indicatedthat the transgene (c-myc-tagged Runx2) isexpressed and its level was increased inresponse to 8BrcAMP treatment in COS-7 cells.Five lines of mice were generated that carry therat MMP-13 promoter (�148) driving expres-sion of the c-myc tagged Runx2 gene. None ofthe transgenic mice lines had significant visualphenotypic changes compared to the wild-typemice.

In order to determine tissue specific express-ion of c-myc tagged Runx2, total RNA was isola-ted from the bones and soft tissue of the 14 daysand 6 weeks old wild-type and transgenic mice.Semi quantitative RT-PCR was carried out todetermine the level of expression of the trans-gene (Fig. 4). In 14 days and 6 weeks oldtransgenic mice, MMP-13 promoter-directedexpression of c-myc-Runx2 was seen in highlymineralized tissues such as tibiae, calvariaeand teeth, whereas in soft tissue (liver) thereappears to be no expression of c-myc-Runx2.There was also expression of the transgene inskin of the 6 weeks old mice. Expression of thetransgene was not detected in either the bone orsoft tissues of wild-type mice. Thus, theseresults indicate that c-myc-Runx2 is expressedonly in transgenic mice and its expression underthe control of the MMP-13 promoter seems to berestricted to bone, teeth, and skin.

MMP-13 Promoter and Runx2 in Mice 5

Author Proof

AFig. 2. Photomicrographs of �456lacZ, �148lacZ, �148(A3R3)lacZ transgenicmice, and wild-type mice. Bones and tissues from14, 16, and 17 day postnatal mice, and 17.5 day prenatal mouseembryos were analyzed immunohistochemically for E. colib-galactosidase. A: 14 day postnatal �148lacZ mouse femur (i),14 day postnatal �148(A3R3)lacZ mouse femur (ii), and 14 daypostnatal wild-type mouse femur (iii), 14 day postnatal�456lacZmouse tail (iv), 14 day postnatal �148(A3R3)lacZ mouse tail (v),and 14 day postnatal wild-type mouse tail (vi), 16 day postnatal�148lacZ mouse calvaria (vii), 16 day postnatal�148(A3R3)lacZmouse tail (viii), and 16 day postnatal wild-type mouse tail (ix),

17 day postnatal �456lacZ mouse calvaria (x), 17 day postnatal�148lacZ mouse calvaria (xi), and 17 day postnatal �148(A3R3)lacZ mouse calvaria (xii). B: 17 day postnatal �148lacZmouse femur (i), 17 day postnatal �456lacZ mouse femur (ii),and 17 day postnatal �148(A3R3)lacZ mouse femur (iii), 17 daypostnatal �456lacZ mouse rib (iv), and 17 day postnatal �148(A3R3)lacZ mouse rib (v), E17.5 day wild-type mouse vertebra(vi), E17.5 day �148lacZ mouse vertebra (vii), and E17.5 day�148(A3R3)lacZ mouse vertebra (viii), 14 day postnatal�148lacZ mouse heart (ix), kidney (x), liver (xi), and lung (xii).

COLOR

Author Proof

AThe wild-type and transgenic mice overex-

pressing Runx2 transcription factor under thecontrol of the MMP-13 promoter were examinedfor changes in their bone phenotype. The pro-ximal tibiae of wild-type and transgenic mice(14 days and 6 weeks) were analyzed by micro-computed tomography (microCT) measure-ments. There were no statistically significantdifferences between the wild-type and trans-genic animals. However, the mCT analysis of theproximal tibiae of the 6 weeks old animalssuggested a tendency for the transgenic animalsto have slightly greater trabecular bone thanthe wild-type (Table IA,B). Histological sectionsof the midtibial metaphysis from wild-type andtransgenic mice (6 weeks old) showed that thereis increased size and number of trabeculae intransgenic mice, compared to wild-type mice(Fig. 5).

To further understand if the slightly increa-sed bone in 6 weeks old transgenic mice is dueto increased bone formation or reduced boneresorption, kinetic analyses of bone formationwere performed. Calcein was injected twice at12 and 2 days before sacrifice in 6 weeks oldmice. Bone histomorphometric studies werecarried out with proximal tibial metaphyses,a standard site for this study. As shown byhistomorphometric data (Fig. 6), the percentageof mineralizing surface, the amount of secretedand mineralized matrix per osteoblast (i.e., themineral apposition rate), and the rate of boneformation were significantly increased in trans-genic mice. Moreover, there was no change inthe number of osteoblasts but the number ofosteoclasts was significantly decreased in thetransgenic mice.

In order to evaluate the molecular events thatunderlie the slightly modified phenotype intransgenic mice, we analyzed the mRNAexpression patterns of genes that are involvedin bone formation (ALP, OC, OPN) and boneresorption (MMP-13, OPG, RANKL). TotalRNA was obtained from tibiae, calvariae, teeth,skin, and liver and was subjected to real time(quantitative) RT-PCR. In the immature skele-ton (14 days old animals), mRNA expression ofbone formation genes such as OC and OPN weresignificantly increased in the tibiae of trans-genic mice (Fig. 7A) and had returned to normalin 6 weeks old transgenic mice (Fig. 7B). Incalvariae, there was increased mRNA expres-sion of OC in both 14 days and 6 weeks oldtransgenic mice (Figs. 7A,B) and there was

Fig. 3. Generation of transgenic mice overexpressing Runx2. A:Structure and restriction map analysis of the�148/c-myc/Runx2.The pCMV-myc mammalian expression vector expressing the N-terminal c-myc epitope tag was used to clone the rat MMP-13promoter (�148 containing the AP-1 and RD sites) and Runx2(type II). The entire fragment containing the rat MMP-13promoter, c-myc epitope tag, and Runx2 was released by aSph1 and PvuII digestion and used for injection into mouseblastocysts. B: Expression of the transgene in vitro. The transgenicDNA construct (p-148/c-myc/Runx2) as well as control con-structs (p-148 and pCMV/c-myc) were transiently transfectedinto COS-7 cells using Lipofectamine reagent for 48 h and thentreated with or without 8BrcAMP (10�3 M) for 24 h. The cellswere then lysed and the c-myc-tagged Runx2 was identified byWestern blot using the monoclonal antibody to the c-mycepitope tag.

Fig. 4. Semiquantitative RT-PCR analysis of the temporalexpression of the transgene in transgenic mice. Total RNA wasextracted from tibiae, skin, teeth, calvariae, and liver of wild-typeand transgenic mice at the indicated ages. One step RT-PCR wascarried out using an RT-PCR kit (Invitrogen) with the forward (c-myc) and the reverse (Runx2) primers. To normalize the amountsof RNA used in the experiment,b-actin was included as a control.The products were identified on a 2% agarose gel.

MMP-13 Promoter and Runx2 in Mice 7

Author Proof

Aincreased expression of OPG and RANKLmRNAs in 6 weeks old transgenic mice(Fig. 7B). In teeth, OPG and RANKL mRNAswere increased in 14 days old transgenic mice(Fig. 7A); whereas expression of OC and MMP-13 mRNAs were increased in this tissue in6 weeks old transgenic mice (Fig. 7B). In skin,mRNA expression of the bone forming genesand bone resorbing genes was not significantlyaltered in either 14 days or 6 weeks old trans-genic mice (Figs. 7A,B). Interestingly, MMP-13mRNA expression was significantly reducedin the tibiae and calvariae of both 14 daysand 6 weeks old transgenic mice (Figs. 7A,B).

When we examined the endogenous Runx2mRNA levels, there was no significant changebetween the wild-type and the transgenicmice.

DISCUSSION

Our previous work has determined the ele-ments and proteins regulating the MMP-13promoter in PTH-treated and differentiatingosteoblastic cells in vitro [Selvamurugan et al.,1998; Winchester et al., 2000]. In both cases, theAP-1 and Runx/RD/Cbfa sites are involved.However, this does not mean that these ele-ments are the functional elements directingosteoblast-preferential expression of this genein vivo. While transient and stable transfectionstudies can provide insight into the biochemicaland molecular interaction, it is only within theintact tissue that the true biological importanceof the promoter and transcriptional environ-ment can be appreciated. This is the rationalefor the extensive and expensive effort requiredto carry out a promoter analysis in intact mice.Transgenic mice provide an excellent settingfor studying complex regulatory systems thatcannot be modeled in isolated experimentalsystems in vitro or in cellulo.

In this study, we show that wild-type trans-genic lines (�456 and �148) express b-galacto-sidase expression in bone, teeth, and skin andnone in heart, liver, or lung, compared to themutant and non-transgenic lines (Fig. 2). Since

TABLE IA. Structural Parameters of Trabecular Bone in the ProximalTibia of 6 Week Old Mice Measured by Micro-Computed Tomography

Parameter Wild-type Transgenic

Percent bone volume, BV/TV (%) 22.33� 0.036 23.14� 0.016Trabecular number, Tb.N (1/mm) 6.22� 0.361 6.48� 0.374Trabecular thickness, Tb.Th (mm) 48.9� 0.003 48.6� 0.001Trabecular separation, Tb.Sp (mm) 155.18� 0.005 148.74� 0.011Connectivity density, Conn.D (1/mm) 221.99� 29.98 266.88� 25.97

Data was tabulated as the mean�SEM. The number of animals/group was five.

TABLE IB. Structural Parameters of Cortical Bone in the Mid-ShaftFemur of 6 Week Old Mice Measured by Micro-Computed Tomography

Parameter Wild-type Transgenic

Percent bone volume, BV/TV (%) 52.77� 0.87 53.10� 0.44Periosteal perimeter, Ps.Pm (mm) 8.41� 0.19 8.37� 0.10Endocortical perimeter, Ec.Pm (mm) 5.33� 0.23 5.29� 0.33Cortical thickness, Ct.Th (mm) 0.194� 0.006 0.191� 0.001

Data was tabulated as the mean�SEM. The number of animals/group was five.

Fig. 5. Histological appearance of Runx2 transgenic bone.Longitudinal sections through the proximal tibiae of wild-type(WT) and transgenic (TG) mice at 6 weeks of age. The trabecularstructure of both WT and TG tibiae has been magnified.

COLOR

8 Selvamurugan et al.

Author Proof

AFig. 6. HistomorphometricQ4 bone formation and resorptionparameters in Runx2-overexpressing mice. Wild-type andtransgenic female mice were labeled with calcein and sacrificedat 6 weeks of age. A: Number of osteoblasts and (B) numberof osteoclasts/bone surface are compared between wild-type(white bars) and transgenic mice (black bars) at 6 weeks of age.

The analyses were done using proximal parts of tibiae.C: mineralizing surface, (D) mineral apposition rate, and,(E) bone forming rate in trabecular bone of wild-type (whitebars) and transgenic (black bars) mice. Bars show means� SEM(n¼ 4). *Significant difference compared to the wild-type mice;P<0.05.

Fig. 7. Quantitative analysis of expression of genes involvedin bone formation and bone resorption. Total RNA was isolat-ed from tibiae, calvariae, teeth, skin, and liver of 14 d (A), and6 weeks (B) old wild-type and Runx2 transgenic mice andsubjected to real time quantitative RT-PCR using specific primersas outlined in the figure and methods section. The mRNAs werenormalized tob-actin. The mRNA level for each gene in the tibiae

of wild-type mice has been converted to 100% and from this therelative mRNA expression was compared in other tissues of bothwild and transgenic mice. The data are represented as mean�SEM (n¼ 3). The experiment was carried out at least three times.*Significant increase compared to the tissues of the wild-typemice; P<0.05.

MMP-13 Promoter and Runx2 in Mice 9

Author Proof

Aboth �456 and �148 reporter constructs exhib-ited a similar regulatory expression in bone,teeth, and skin, we suggest that the regulatoryelements (AP-1 and Runx/RD/Cbfa) within the148 bp upstream of the MMP-13 promoter areenough to confer this effect. In addition, it isevident that transgenic mice containing muta-ted AP-1 and Runx/RD/Cbfa sites (�148A3R3)in the MMP-13 promoter expressed very lowb-galactosidase expression, compared to thenon-transgenic lines (Fig. 2). Hence, it is reaso-nable to assume that activation of the MMP-13promoter requires only the AP-1 and Runx/RD/Cbfa sites in both in vitro and in vivo conditions.

Runx2 is a major regulator of bone develop-ment [Karsenty, 2000; Komori, 2000]. Mousemodels have enhanced our understanding of thebasic functions of Runx2. Mice heterozygouslymutated in the Runx2 locus show a phenotypesimilar to that of cleidocranial dysplasia (CCD)in humans, in whom mutations of Runx2 havebeen found [Komori et al., 1997; Mundlos et al.,1997; Otto et al., 1997]. A homozygous mutationof this gene in mice induced a complete lack ofbone formation with arrest of osteoblast differ-entiation [Komori et al., 1997; Otto et al., 1997].The dominant negative form of Runx2 devel-oped an osteopenic phenotype in mice and wasused to indicate the indispensability of the genefor postnatal bone formation by regulating thefunctions of mature osteoblasts [Ducy et al.,1999]. Through deletion of the C-terminal

intranuclear targeting signal by homologousrecombination, it has been shown that sub-nuclear targeting and the associated regulatoryfunctions are essential for control of Runx-dependent genes [Choi et al., 2001]. A largenumber of in vitro studies have also implied thatRunx2 is a positive regulator that can stimulatethe expression of bone matrix genes, includingtype I collagen, osteopontin, bone sialoprotein,osteocalcin, and fibronectin [Banerjee et al.,1997; Ducy et al., 1997; Sato et al., 1998; Haradaet al., 1999; Xiao et al., 1999; Lee et al., 2000;Kern et al., 2001; Prince et al., 2001].

A fundamental tool that is used in thetransgenic experimental approach is a promo-ter that has tissue-restricted activity. Withinthe lineage of bone and cartilage cells, the type Iand type II collagen promoters can be designedto have preferential expression at specificstages of differentiation. OC and BSP expres-sion is specific to bones and thrombocytes andthe OC promoter has been widely used in thetransgenic mouse model system [Clark andRowe, 2002]. When the MMP-13 promoter wasused to overexpress the b-galactosidase repor-ter gene, we found its expression not only inbone but also in teeth, and skin (Fig. 2).

Liu et al. [2001] reported that transgenic miceexpressing Runx2 directed by the pro-a-type Icollagen promoter had osteopenia and fragilityof bone that were caused by the inhibition of os-teoblast maturation, and immature osteoblasts

Fig. 7. (Continued )

10 Selvamurugan et al.

Author Proof

Aaccumulated in the bone of adult mice [Liu et al.,2001]. Their transgenic mice showed decreasesin bone formation rate, matrix apposition rate,and mineralized surface area in trabecular boneas well as in cortical bone compared to those ofthe wild-type mice. Geoffroy et al. [2002] havereported that Runx2 controls not only genesthat are important for osteoblast differentiation[Ducy et al., 1997] and function [Ducy et al.,1999] but also genes that are involved inosteoclast differentiation and bone formation-resorption coupling [Geoffroy et al., 2002].Even though MMP-13 has been shown to beexpressed in skin, uterus, and ovary, it is mostlyexpressed in bone [Balbin et al., 1996; Daviset al., 1998; Tuckermann et al., 2000; Shumet al., 2002]. We report here that overexpressionof Runx2 directed by the MMP-13 promoterincreases the bone mineralization surface, boneformation rate, and matrix apposition rate(Fig. 6). Since there was no change in the num-ber of osteoblasts, this effect could be due touncoupling and unbalancing of bone formationand bone resorption processes. Even thoughOPG and RANKL mRNA expression were notaltered in the tibiae of transgenic mice at theages examined, there was significantly reducedMMP-13 mRNA expression during develop-ment of transgenic mice (Fig. 7A,B) and it hasbeen shown that MMP-13 is necessary for osteo-clast-mediated bone resorption [Zhao et al.,2000]. The downregulation of MMP-13 at theages examined by overexpression of Runx2 intransgenic mice could be due to negative feedback regulation of Runx2. This also could be dueto the fact that MMP-13 is expressed at greaterlevels in long bones, in the fetus, and maximallyat 14 days in the calvariae, while OC is mostlyexpressed postnatally [Davis et al., 1998;Tuckermann et al., 2000]. It is possible thatthe maturation of the bones is advanced andthe usual peak in MMP-13 expression is at anearlier age. Perhaps the reduced expressionlevel of MMP-13 in the transgenic mice couldhave led to decreased recruitment of osteoclasts(Fig. 6) to the bone surface, resulting in reducedbone-resorptive activity, reflected by increasedbone formation in transgenic mice.

Overall, we provide evidence that the 148base pairs of MMP-13 promoter is sufficientand necessary for tissue-restricted (bone, teeth,and skin) gene expression in vivo, and theAP-1 and Runx/RD/Cbfa sites are likely to regu-late this. Using these regulatory elements, we

further document that overexpression of Runx2appears to alter the balance between the boneformation-bone resorption processes in vivo anddoes regulate the expression of MMP-13 andother bone marker genes.

ACKNOWLEDGMENTS

We thank Dr. Gerard Karsenty for the Runx2construct that was used to generate the variousRunx2 vectors. We also thank Diana Galperina,Ziawei Fung, and Olivia Linton for their tech-nical assistance.

REFERENCES

Aubin JE. 1998. XXXQ1. J Cell Biochem 30–31(Suppl):73–82.

Avioli LV, Krane SM. 1998. Metabolic bone disease, 3rdedition. XXXQ2: Academic Press.

Balbin M, Fueyo A, Lopez JM, Diez-Itza I, Velasco G,Lopez-Otin C. 1996. J Endocrinol 149:405–415.

Banerjee C, McCabe LR, Choi JY, Hiebert SW, Stein JL,Stein GS, Lian JB. 1997. J Cell Biochem 66:1–8.

Choi JY, Pratap J, Javed A, Zaidi SK, Xing L, Balint E,Dalamangas S, Boyce B, van Wijnen AJ, Lian JB, SteinJL, Jones SN, Stein GS. 2001. Proc Natl Acad Sci USA98:8650–8655.

Clark S, Rowe D. 2002. Principles Bone Biol XXQ3:xxx–xxx.D’Alonzo RC, Selvamurugan N, Karsenty G, Partridge NC.

2002. J Biol Chem 277:816–822.Davis BA, Sipe B, Gershan LA, Fiacco GJ, Lorenz TC,

Jeffrey JJ, Partridge NC. 1998. Calcif Tissue Int 63:416–422.

Ducy P, Zhang R, Geoffroy V, Ridall AL, Karsenty G. 1997.Cell 89:747–754.

Ducy P, Starbuck M, Priemel M, Shen J, Pinero G, GeoffroyV, Amling M, Karsenty G. 1999. Genes Dev 13:1025–1036.

Geoffroy V, Kneissel M, Fournier B, Boyde A, Matthias P.2002. Mol Cell Biol 22:6222–6233.

Harada H, Tagashira S, Fujiwara M, Ogawa S, KatsumataT, Yamaguchi A, Komori T, Nakatsuka M. 1999. J BiolChem 274:6972–6978.

Inada M, Wang Y, Byrne MH, Rahman MU, Miyaura C,Lopez-Otin C, Krane SM. 2004. Proc Natl Acad Sci USA101:17192–17197.

Karsenty G. 1999. Genes Dev 13:3037–3051.Karsenty G. 2000. Semin Cell Dev Biol 11:343–346.Kern B, Shen J, Starbuck M, Karsenty G. 2001. J Biol

Chem 276:7101–7107.Komori T. 2000. Biochem Biophys Res Commun 276:

813–816.Komori T, Kishimoto T. 1998. Curr Opin Genet Dev 8:

494–499.Komori T, Yagi H, Nomura S, Yamaguchi A, Sasaki K,

Deguchi K, Shimizu Y, Bronson RT, Gao YH, Inada M,Sato M, Okamoto R, Kitamura Y, Yoshiki S, Kishimoto T.1997. Cell 89:755–764.

Kusano K, Miyaura C, Inada M, Tamura T, Ito A, NagaseH, Kamoi K, Suda T. 1998. Endocrinology 139:1338–1345.

MMP-13 Promoter and Runx2 in Mice 11

Author Proof

ALee KS, Kim HJ, Li QL, Chi XZ, Ueta C, Komori T, Wozney

JM, Kim EG, Choi JY, Ryoo HM, Bae SC. 2000. Mol CellBiol 20:8783–8792.

Liu W, Toyosawa S, Furuichi T, Kanatani N, Yoshida C, LiuY, Himeno M, Narai S, Yamaguchi A, Komori T. 2001.J Cell Biol 155:157–166.

Mundlos S, Otto F, Mundlos C, Mulliken JB, Aylsworth AS,Albright S, Lindhout D, Cole WG, Henn W, Knoll JH,Owen MJ, Mertelsmann R, Zabel BU, Olsen BR. 1997.Cell 89:773–779.

Otto F, Thornell AP, Crompton T, Denzel A, Gilmour KC,Rosewell IR, Stamp GW, Beddington RS, Mundlos S,Olsen BR, Selby PB, Owen MJ. 1997. Cell 89:765–771.

Parfitt AM, Drezner MK, Glorieux FH, Kanis JA, MallucheH, Meunier PJ, Ott SM, Recker RR. 1987. J Bone MinerRes 2:595–610.

Partridge NC, Davis BA, Sipe B, Gershan LA, Fiacco GJ,Lorenz TC, Jeffrey JJ. 1998. Calcif Tissue Int 63:416–422.

Prince M, Banerjee C, Javed A, Green J, Lian JB, Stein GS,Bodine PV, Komm BS. 2001. J Cell Biochem 80:424–440.

Roodman GD. 1999. Exp Hematol 27:1229–1241.Sato M, Morii E, Komori T, Kawahata H, Sugimoto M,

Terai K, Shimizu H, Yasui T, Ogihara H, Yasui N, Ochi T,Kitamura Y, Ito Y, Nomura S. 1998. Oncogene 17:1517–1525.

Schorpp M, Mattei MG, Herr I, Gack S, Schaper J, Angel P.1995. Biochem J 308:211–217.

Scott DK, Brakenhoff KD, Clohisy JC, Quinn CO, PartridgeNCIPNC. 1992. Mol Endocrinol 6:2153–2159.

Selvamurugan N, Chou WY, Pearman AT, PulumatiMR, Partridge NC. 1998. J Biol Chem 273:10647–10657.

Shum JK, Melendez JA, Jeffrey JJ. 2002. J Biol Chem 277:42830–42840.

Tuckermann JP, Pittois K, Partridge NC, Merregaert J,Angel P. 2000. J Bone Miner Res 15:1257–1265.

Varghese S, Ramsby ML, Jeffrey JJ, Canalis E. 1995.Endocrinology 136:2156–2162.

Varghese S, Delany AM, Liang L, Gabbitas B, Jeffrey JJ,Canalis E. 1996. Endocrinology 137:431–437.

Varghese S, Rydziel S, Canalis E. 2000. Endocrinology141:2185–2191.

Winchester SK, Selvamurugan N, D’Alonzo RC, PartridgeNC. 2000. J Biol Chem 275:23310–23318.

Xiao ZS, Hinson TK, Quarles LD. 1999. J Cell Biochem74:596–605.

Zhao W, Byrne MH, Boyce BF, Krane SM. 1999. J ClinInvest 103:517–524.

Zhao W, Byrne MH, Wang Y, Krane SM. 2000. J Clin Invest106:941–949.

Q1: Please provide article title for all journalreferences.

Q2: Please provide publisher location.

Q3: Please provide volume number and page range.

Q4: Please provide better quality of artworks.

12 Selvamurugan et al.